Options Scorecard: The Week of August 3, Graded (37 Calls, 59% Right)

TL;DR

• Every options play logged across the week of August 3, from Palantir's Monday print through Sea Limited's Tuesday reaction, scored against what actually happened: 22 of 37 resolved calls were right, a 59% hit rate.
• The biggest single miss was passing on Palantir's straddle. The stock ran 29.4% against a 10-15% implied move logged at $125.65, and this page took neither side of that trade.
• One call gets reversed here. Super Micro's straddle pass was graded a win in real time off an early 6-9% after-hours read. The stock kept running to a 32.4% close-to-close gain against an 18% implied move. The early grade was wrong, and this scorecard corrects it to a loss.
• The week split roughly down the middle on calibration. Five names (AMD, Take-Two, Uber, Sea Limited, and Datadog's short-premium leg) rewarded selling volatility; seven blew through their implied moves entirely. Neither reflex was right on its own this week.
• Datadog and The Trade Desk were this page's only two genuine buy calls this week, and both won big: Datadog's straddle cleared a 13% implied on a ~17% realized move, and The Trade Desk's cleared a 13% implied on a 24.6% drop.

The Board

22 of 37 right. The single worst call was standing aside on Palantir. The single most instructive one was catching our own bad grade on Super Micro.

This is the site's first scorecard, so it covers the whole backlog: every trade log from the week-ahead hub of August 3-7 plus the handful of previews that spilled into the following week and have since resolved. Going forward it runs against the current week-ahead hub and whatever that week's previews log.

The Losses, Named

Fifteen calls lost this week. Grouped by mechanism, because a list of tickers teaches nothing on its own.

Passed on the move, and it blew through implied (seven calls)

This is the July pattern repeating: standing neutral, or betting small, into a name that had a real catalyst.

• Palantir (Row 1, pass on a long straddle). Logged against a $125.65 Monday close and a 10-15% implied move. The stock closed Tuesday at $162.66, up 29.4%, after revenue grew 93% and the full-year guide jumped to $8.15-8.16 billion. Buying the straddle would have paid for itself more than twice over. This is the week's biggest single miss.
• Shopify (Row 1, pass on a long straddle). Logged at the site's own corrected $126.88 Monday close against a ~12.5% implied move. Wednesday's close of $144.24 was +13.7%, clearing the implied even on the conservative reading.
• SpaceX (the Aug 4 straddle pass, carried across the preview and the lockup piece). Logged at $108.37 against a 12.5% weekly implied. Friday's close of $133.11 was +22.8%, through the unlock and the first earnings print at once.
• Super Micro (Row 1, pass on a long straddle). See the correction below. Logged at $28.40 against an 18% implied; the real close-to-close move was +32.4%.
• Oklo (pass on any pre-print position). Logged at $41.22 against a ~13% implied. The Groves reactor going critical Wednesday only moved the stock about 4% that day; the real move came Friday alongside the earnings print and a reactor-sales update, closing at $48.42, +17.5% for the week. The original write-up blamed the wrong day and the wrong size for this one; corrected here.
• AppLovin (Row 1, pass on a long straddle). Logged at $406.16 against a 12-13% implied. Thursday's regular close of $335.67 was -17.4%, worse than the after-hours print this page originally quoted.
• SK Hynix (both rows, long calls and short puts). No implied move was ever logged here, so this one is graded on direction alone: the ADR ran from $143.73 to $165.67, +15.3%, while the shareholder-return announcement the piece was actually watching for still has not been made. Two passes, one direction, two losses.

Passed on selling premium, and the move stayed inside implied (six calls, one approximate)

The mirror image, and the more surprising half of the week: several names that usually overshoot did not.

• AMD (Row 2, pass on selling premium). Realized move from the $484.64 spot to Friday's close was essentially flat, well inside the 12.3% implied. Selling premium here would have been the easy trade.
• Take-Two (Row 2, pass on selling premium). Friday's close of $246.50 was +1.5% from the $242.92 spot (+6% from the last pre-print close), against a 7.7% implied. Same story.
• Uber (Row 2, pass on selling premium). Wednesday's close of $68.18 was -4.8% against a 7-8% implied, even with gross bookings of $58 billion clearing the top of guidance.
• Sea Limited (Row 2, pass on selling premium). Realized +14.5% against a 19.1% implied. A premium seller collected clean.
• DraftKings (Row 2, pass on selling premium). The earnings-day move was +8.4% against a ~9% implied, a near miss that still favored the seller.
• Meta (Row 3, pass on selling premium). Logged against an approximate ~$580 intraday quote on the Iran-driven Monday gap. The follow-through by Friday was only around 2%, small enough that a premium seller would have collected. The baseline here is a rough intraday read rather than a close, so treat the margin as approximate, not the direction.

The one structural loss

• Google DeepMind (Row 3, short premium in a $366-381 strike zone). The stock closed the week at $354.30 and kept falling to $343.80 by August 11, a full 9.5% below the zone the position was written around. A defined range only protects the seller while price stays inside it.

The Correction: Super Micro

Formula matters here as much as the result. Super Micro's straddle pass was marked a win in the original preview less than an hour after the print, using the first after-hours reaction of roughly 6-9%, which sat comfortably inside the 18% implied move. That was the only number available at the time.

It was also the wrong number to stop on. The stock kept climbing through Wednesday's session and closed at $37.61, a 32.4% move from the $28.40 spot. Grading a straddle off the first tick of an after-hours print instead of the settled close is exactly the mistake this site's own trade-log rules exist to prevent, and it is corrected here: Super Micro's straddle pass is a loss, not a win. The premium-selling pass on the same name is still a win, since a seller at 18% implied against a 32% realized move would have been the one blown out, and this page never took that side either.

What Won

The rest of the ledger, briefer, because wins need less explaining than losses do.

Two real buy calls, both paid. Datadog's long straddle cleared its 13% implied on a realized move near 17%, the second straight quarter this exact setup rewarded buying volatility. The Trade Desk's straddle cleared its 13% implied on a 24.6% collapse after a soft Q3 guide.

Correctly avoided selling into a gap: Palantir Row 2, Shopify Row 2, Super Micro Row 2, and the SpaceX put pass logged against $108.37, all of which would have been blown through by the moves above.

Correctly stayed out of a straddle that would have lost: AMD Row 1, Take-Two Row 1, Uber Row 1, Sea Limited Row 1, DraftKings Row 1, Vistra, Disney, Eli Lilly, Cava (corrected below), and Micron Row 1's puts, which would have expired worthless against a stock that rallied 9.6% into Friday instead of falling further.

Cava's own grade needed the same fix as Super Micro's, in the other direction. The original preview marked its pass a loss off a +15.3% intraday quote captured Wednesday morning. The regular-session close-to-close move, from Tuesday's $60.81 pre-print close to Wednesday's $69.47, was +14.2% against an implied logged at 11-14%, a near miss at worst, and from the $65.23 spot originally logged it was just +6.5%, comfortably inside implied either way. Corrected here to a win.

GOOGL Row 1 and Meta Row 2, both passes on chasing calls into a move that never extended far enough to cover the premium, also come in as wins.

Five Positions Still Open

These triggered and have a real mark, but haven't closed, so they sit outside the hit rate above rather than padding it.

Ticker	Entry	Latest mark	Unrealized
SpaceX shares	$114.92, Aug 6 close	$141.29, Aug 13 close	+23.0%
Super Micro shares	$34.99, Aug 12 open	$39.16, Aug 13 close	+11.9%
Sea Limited shares	$114.91, entry close	$128.11, Aug 12 close	+11.5%
SanDisk confirmation calls	Struck near $1,258.58, Aug 6 open	Underlying $1,528.11, Aug 13 close	Underlying +21.4%
AMD conditional long	~$482.05 proxy, Aug 5	$483.36, Aug 7 close	+0.3%, roughly flat

Six more conditions triggered this week with no sourced entry price to mark: Palantir Row 3, Shopify Row 3, Disney Row 2, Eli Lilly Row 2, Airbnb Row 2, and Circle Row 2. They carry to next week's scorecard once an entry price is sourced rather than getting an invented one now.

The Calibration Lesson

July's lesson was simple: realized moves kept beating implied all season, and the reflex to sell an "expensive-looking" implied move lost money over and over. This week does not repeat that cleanly. Seven calls lost because a move blew through implied and five lost because a move stayed inside it, close to an even split.

The pattern underneath the split is more useful than the split itself. Every name that blew through its implied move this week had a genuine surprise attached: a guide raised past the top of its own range (Palantir), a reactor confirmation plus a sales update landing days apart (Oklo), an unlock and a first earnings print stacked on the same week (SpaceX), a print that kept running well past the first tick (Super Micro). Every name that stayed inside implied had already told the market what to expect: AMD's data-center number sat inside its own guide, Vistra and Disney both reiterated, Uber's bookings beat but the guide had signaled it. The lesson for next week is not "buy volatility" or "sell volatility." It is that a name confirming a number the market already has is a premium-selling opportunity, and a name that changes the number is not, and the confusion this week came from applying last month's blanket instinct to setups that needed the distinction.

The Full Ledger

Every scored row, in one place.

Ticker	Play	Logged	Result	Realized vs. implied
PLTR	Pass, long straddle	$125.65, 10-15% implied	Loss	+29.4% vs. implied
PLTR	Pass, short premium	same	Win	+29.4% vs. implied
AMD	Pass, long straddle	$484.64, 12.3% implied	Win	~flat vs. implied
AMD	Pass, short premium	same	Loss	~flat vs. implied
SHOP	Pass, long straddle	$126.88, 12.5% implied	Loss	+13.7% vs. implied
SHOP	Pass, short premium	same	Win	+13.7% vs. implied
DDOG	Buy, long straddle	$273.60, 13% implied	Win	~17% vs. implied
DDOG	Pass, short premium	same	Win	~17% vs. implied
TTD	Buy, long straddle	$18.30, 13% implied	Win	-24.6% vs. implied
TTD	Pass, short premium	same	Win	-24.6% vs. implied
TTWO	Pass, long straddle	$242.92, 7.7% implied	Win	+1.5% vs. implied
TTWO	Pass, short premium	same	Loss	+1.5% vs. implied
VST	Pass, any position	$141.38 pre-print, 6.4% implied	Win	-0.6% vs. implied
OKLO	Pass, any position	$41.22, ~13% implied	Loss	+17.5% vs. implied
CAVA	Pass, any position	$65.23, 11-14% implied	Win	+6.5% to +14.2% vs. implied
UBER	Pass, long straddle	$71.61, 7-8% implied	Win	-4.8% vs. implied
UBER	Pass, short premium	same	Loss	-4.8% vs. implied
DIS	Pass, any position	$98.14, ~6% implied	Win	+3.7% vs. implied
LLY	Pass, any position	$1,121.36, ~7% implied	Win	~+4.5% vs. implied
SPCX	Pass, long put	$108.37, 12.5% implied	Win	+22.8% vs. implied
SPCX	Pass, long straddle	same	Loss	+22.8% vs. implied
SMCI	Pass, long straddle	$28.40, 18% implied	Loss (corrected)	+32.4% vs. implied
SMCI	Pass, short premium	same	Win	+32.4% vs. implied
SE	Pass, long straddle	$114.91, 19.1% implied	Win	+14.5% vs. implied
SE	Pass, short premium	same	Loss	+14.5% vs. implied
MU	Pass, long puts	$801 premarket, directional	Win	+9.6%, wrong direction for puts
GOOGL	Pass, chasing calls	$379.72 intraday	Win	-9.5% by Aug 11
GOOGL	Short premium, $366-381 zone	same	Loss	Range breached to $343.80
META	Pass, chasing calls	~$580 intraday, approximate	Win	~+2%, approximate
META	Pass, short premium	same	Loss	~+2%, approximate
DKNG	Pass, long straddle	$22.17 pre-print, ~9% implied	Win, narrow	+8.4% vs. implied
DKNG	Pass, short premium	same	Loss, narrow	+8.4% vs. implied
APP	Pass, long straddle	$406.16, 12-13% implied	Loss	-17.4% vs. implied
APP	Pass, short premium	same	Win	-17.4% vs. implied
SKHY	Pass, long calls	$143.73, no implied logged	Loss	+15.3%, direction only
SKHY	Pass, short puts	same	Loss	+15.3%, direction only
SNDK	Pass, buy the after-hours dip	$1,350.50, graded vs. Aug 6 close	Win	Aug 6 close $1,258.58, below entry zone

Nine rows could not be scored for lack of a clean logged implied move, a specified strike, or two agreeing sources, and are left out of the hit rate rather than forced: Circle Row 1 (no implied number, only IV30), Airbnb Row 1 (no implied logged), the two generic "any August expiry" premium-selling rows on Micron across both Micron pieces, the SanDisk strangle threshold call whose logged implied move was disputed between 17.5% and 25% with no way to resolve it, SanDisk's own generic premium-selling row, DraftKings Row 3 (the site's own coverage still calls the outcome unverified), and Applied Materials, which reported Thursday evening into after-hours-only data as this piece went to press.

Two Figures Worth Flagging

Datadog's trade log carried a stale $254.79 spot that never matched Monday's actual $273.60 close, a leftover from before the piece's own later update settled on a 17% move. Fixed in place in the original article; the correction doesn't change the win.

Vistra's and Oklo's original trade logs dated their spot price "Aug 3" when the dollar figures logged were actually the prior session's close. Both are corrected in this scorecard's ledger; neither correction flips the verdict.

The One-Line Read

Fifty-nine percent is a real number from a full ledger, and its two most important entries are both losses: a Palantir straddle nobody bought, and a Super Micro grade this page got wrong and is fixing now, in public.