THE MECHANISM DESIGN

Game theory run backwards: fix the rules so that honesty is the best move. The engine is the second-price (Vickrey) sealed-bid auction — highest bidder wins and pays the second-highest bid. Truth-telling becomes a dominant strategy, proven exactly over a grid of every deviation.

source W. Vickrey, Counterspeculation, Auctions, and Competitive Sealed Tenders, Journal of Finance 16(1), 1961, pp. 8–37 · doi.org/10.1111/j.1540-6261.1961.tb02789.x · rendered, not quoted.

Blue Team · builds & defends
3
The Model

Each bidder k holds a private value vk and submits a sealed bid bk. The mechanism is a pair of functions on the bid vector:

winner = argmaxk bk
price = maxj≠winner bj (2nd-highest bid)

Utility is uk = vk − price if k wins, else 0. Because the price a winner pays depends only on others' bids, no bidder can lower their own price by shading — so reporting vk truthfully is a dominant strategy.

5
The Lineage

Neighbour sphere: THE NASH EQUILIBRIUM — designed on purpose. Nash asks "given the rules, where do rational agents settle?" Mechanism design inverts it: choose the payoffs so the equilibrium is truth-telling.

Vickrey does better than a Nash equilibrium: truthful bidding is dominant (best whatever others do), not merely a mutual best-response. Strategyproofness — the corner of game theory that alignment and governance build on.

7
The Witness

Live re-check of the running engine: it searches every bidder, value and deviation on the grid for a profitable underbid. Truthful ⇒ none exists.

booting…

Flip the payment rule in window 6 and this witness turns red within one render.

The Machine
4
Data In in ↓

Three bidders with fixed private values. Bidders 2 & 3 bid truthfully; you drive bidder 1's bid to test whether deviating ever helps.

bidder 1
bidder 2
bidder 3

true values  v = (7, 4, 2) · truthful bid = 7

0
The Panel LIT

Live outcome under the current rule (second-price):

allocation
price paid by winner
bidder 1 utility (v₁ − price)
best bidder-1 utility attainable

8
Data Out out ↓

Proven result — winner is the highest-value bidder, price is the second-highest value, and truth-telling is dominant:

booting…
Red Team · attacks & breaks
1
The Adversary WALL
"I will lie about my value to pay less." Under Vickrey this is self-defeating: shade below your value and you only risk losing an item you would profit from — the price you would pay is untouched. Overbid and you risk paying more than the item is worth. The only bid that never regrets is vk.

The adversary's real leverage is elsewhere: collusion (rings agreeing to suppress bids) and a corrupt/uncommitted auctioneer who peeks at bids or invents a phantom second bid. Vickrey is truthful only when the seller is bound to the rule.

2
The Graveyard

"Second-price just means the winner pays less, so sellers lose money."
Correction: revenue equivalence — in expectation, first- and second-price yield the same revenue with symmetric bidders; Vickrey buys truthfulness, not lower revenue.

"Truthful ⇒ also efficient in revenue for the seller."
Correction: Vickrey maximises allocative efficiency, not seller revenue. amber The revenue-maximising design (Myerson 1981) sets reserve prices and generally is not the plain Vickrey rule.

"Dominant-strategy truthful ⇒ safe to run in practice."
Correction: breaks under budget constraints, common values (winner's curse), and multi-unit settings.

6
The Tamper

Disclosed planted void: switch the rule to charge the winner their own bid (turn second-price into first-price). Truthful bidding is no longer dominant — the witness in window 7 catches a profitable underbid.

Under first-price a winner with value 5 facing a bid of 2 does strictly better bidding 3 (u = 2) than bidding 5 (u = 0).