Still a draft, not the rules yet

Behind in April
shouldn't mean done by May.

A proposal for next season, not the rules yet. One real idea, tested properly (including a couple of wrong assumptions we caught and fixed along the way), and a straight question: does this feel right to you?

Why bring this up at all

Here's the situation as I see it.

Every season, everyone predicts the full order F1's constructors will finish in, before a single race happens. You earn points for accuracy every race weekend after that, stacked on top of the real points those teams actually score on track. There's a catch, though. You get exactly one chance, ever, for the whole season, to change your original prediction if it's clearly wrong.

That's worked fine when everyone's guess holds up reasonably well. But when someone's prediction is off early, and they've either already used their one swap or never found the right moment to use it, there's no way back. The standings can feel decided by May, with five more months of racing left to sit through. I don't think that's good for the league. I want every season to stay a real race, for whoever's behind just as much as whoever's ahead. That means a real chance to catch up, not free points for falling behind.

First, the reassurance

The game you already love isn't going anywhere.

You still call the full constructor order before lights out. Nailing it, or getting close, is still what actually wins a season. This isn't turning into a driver draft or a salary-cap fantasy team. It's still your league. It just gets a steering wheel instead of one static guess, and a few ways to keep it interesting when someone falls behind.

What's actually on the table

The idea we landed on, and what didn't make it.

We tested a lot more than what's below, and most of it didn't survive contact with real season data. One other idea actually passed every safety check: a scoring tweak that quietly boosted your bonus points the further behind you were. It worked. The numbers were clean. It still didn't make the cut, not because it failed, but because of how it felt to actually play with: it moved your score around based on where you stood in the standings, not on a call you made.

The points budget below is the one we're proposing instead. It stays tied to your own read on the constructors the whole way through; spend it well and you're rewarded for being right, spend it badly and you're not. Paired with two smaller changes below it.

The mechanic

Meet the Trade Budget.

Your one-time swap goes away under this option, replaced by this instead. Rather than one all-or-nothing reset, you get a running points budget to spend correcting individual teams, all season long. No more single free do-over if your whole prediction goes wrong early. You fix it piece by piece instead, same as everything else here.

Everyone starts the season with the same 300 points, free, all available right away. Drag a team to a new predicted spot whenever the trade window's open. The price is as simple as it gets: 5 points for every spot you move a team, the same rate no matter which team it is or where it sits in the order. Move a team 3 spots, that's 15 points. No tiers, no lookup table, just how far you moved it, times 5. You can fix several teams in the same sitting if you want; they're priced together as one trade, not charged one at a time.

The catch: your next trade after that costs more than the last, whether it's five minutes later or five months later. One well-considered trip to the pit wall is cheap. Constantly tinkering gets expensive fast, on purpose, that's what actually rewards making fewer, better-timed calls instead of fiddling all season.

Season budget: 300 pts, all available from day one Per-trade cap: 150 pts, so one trade can't wipe out the whole season's budget Pricing: flat, 5 pts per spot moved, escalates each time you come back
Why 300, not less? We swept it directly: 200 points narrows the gap by about 166 pts, 300 by about 220, the best of everything we tried by a wide margin. Going bigger doesn't keep helping either, 400 drops back to about 165 and 600 to about 192. 300 is a real, clear peak in the data, not a round number picked for the label.
What's that 150 cap actually for? For a typical correction it never comes into play. We tested the real worst case too: a prediction that's almost completely backwards from day one. Uncapped, fixing that in one sitting could cost as much as 205 points in a single trade. The cap doesn't reduce your total points spent, spreading a big fix across more sittings actually costs a little more once the escalation stacks up, it just guarantees no single decision ever costs more than 150. A predictable ceiling on your worst day, not a cheaper one.

Two smaller changes, alongside it

Your score means more, and shows less clutter.

Bullseyes and Close Calls are both worth noticeably more next season than they are today. Same mechanic, same idea, exact calls and near-misses just pay off harder. It's a real shift in how much precision matters relative to a rough guess, worth knowing on its own, separate from anything about trading.

The second change is just about what you see. Right now your season total includes the real points every constructor actually scores on track, added the same to every player regardless of what you predicted. It doesn't decide anything; everyone gets the identical amount, but it makes your displayed score a much bigger number than the part that's actually competitive. Next season your total would just be your bonus points, nothing else. A smaller number, but every point in it is something you actually earned.

The rule this isn't allowed to break

A real shot, not a free pass.

The whole point of this is opportunity, not a guarantee. Nobody should be able to fall behind and get automatically handed the lead back. So it got checked against a simple question, run across a lot of simulated seasons at every league size from 4 to 10 players: today's game alone, who would have actually finished last? Then, with the budget turned on, how often does that same person still end up winning anyway?

Trade Budget on top of today's game, tested across three real F1 seasons

~7.4%
How often whoever'd actually finish last still won anyway, on average
vs.
10–25%
What pure random chance would predict, depending on group size

The toughest single case came back at 8.7%, still under that group size's own chance level of 10%. A real lever, not a guarantee.

See it happen

One real season, replayed two ways.

Everything on this page so far has been the plan. Here's what it actually does. We took one real F1 season, 2023, and six simulated players with realistic tiered predictions, then replayed it twice with the exact same starting predictions and the exact same races: once exactly as the game works today, once with the Trade Budget and the bigger bonus values both turned on. Same season, same players, only the rule changed.

Today's game (unchanged) + Trade Budget
The gap between 1st and last: 425 points today, 210 with the Trade Budget. This one run lands close to the real average across every trial (within a few points), not a best case.

How consistent is that?

One season is one story. Here's all 180.

The chart above is a single example. These are every season and player group actually tested, grouped by how much the gap changed. Bars to the right mean the gap got narrower.

Gap got wider Trade Budget: gap got narrower
168 of 180 came out narrower, 12 came out wider, across every league size tested (4 to 10 players). Average +214 pts narrower for this player count. This is not a guarantee it holds at every possible group size or every season, just an honest account of what was actually checked.

For anyone who wants the receipts

What we actually tested this against.

We didn't have anywhere near enough real seasons of your actual picks to test this properly, so we built stand-in players instead: thousands of them. Each had their own realistic set of predictions, bunched in tiers the way real fans actually guess, favorites near the top and longshots near the bottom, not a random shuffle. Then we let them play out against three real, complete Formula 1 seasons (2023 through 2025) to see what really happened. We also checked it against different league sizes, from 4 players up to 10, not just a group your exact size, so this isn't tuned to one lucky group.

SetupAvg. change in the 1st–last gapWorst case: how often last place still won anyway
Today's game, unchanged (the real baseline)0% (chance 10–25%)
Today's game + Trade Budget + bigger bonus values+218 pts narrower7.4% avg, 8.7% worst case (chance 10–25%)
A note on how we tested this: it's tempting to test new rules against random, shuffled predictions, but real F1 seasons aren't random. Top teams stay near the top most of the time. Backmarkers stay near the back. Testing against random predictions would overstate how much any of this actually helps, so we tested against realistic tiered predictions instead, across three separate real seasons, not just one. We also made sure the comparison was fair: measured against what the league actually has today, not against a hypothetical version with no swap at all, and against a pricing model that matches how a real trade would actually get charged, bundled corrections and all. The bigger bonus values were checked on their own too: they widen the natural gap by more than a simple point increase would suggest, since exact calls get amplified more than close calls, but the same player tends to win either way (92.5% of the time, checked directly), and the Trade Budget absorbs the change without needing any retuning.

Over to you

Does this feel right?

That's the real question now, not a detail to fine-tune. A points budget that stays tied to your own read on the constructors, paired with bonus points that are worth more and a score that shows only what you actually earned. It's tested, it's safe, and it's the one idea we're standing behind.

And it's a fair answer to say you like the league exactly the way it is. I think this is worth trying, but that's my take, not a decision that's already been made. If the honest reaction is "leave it alone," say that. It counts just as much as any other feedback here.