How we score every call (and why most sites never will)
Every pick on this site — mine, the bot’s, the pundits’ — is stamped with a date, an entry price, and a benchmark, then marked to market daily. The full methodology, published so you can attack it.
The finance internet has a memory problem, and the memory problem is the business model. A pundit makes forty calls a year, mentions the four winners forever, and the archive quietly swallows the rest. This site exists to be the opposite of that, so the scoring methodology is public. Here it is in full — attack it if you can find a hole.
The rules
- Every call is logged the day it is made — a ticker, a date, a direction, and a thesis in one or two sentences. No backdating, ever. If a call is logged late (pundit picks often are), the entry uses the price from the day the call was publicly made, and the source is linked.
- The entry price is the closing price on the call date. Not an intraday cherry-pick. If markets were closed, the next trading day's close.
- Every call is marked against an index, not against zero. Canadian calls are measured against the S&P/TSX Composite; US calls against the S&P 500. Being up 8% while the index is up 12% is a losing call here. Beating the market is the claim every stock picker implicitly makes — so it is the test we apply.
- Marks run automatically every trading day. A systemd job fetches closes and recomputes every call's return versus its index — including the embarrassing ones. Especially the embarrassing ones. There is no human step where a bad number could get quietly skipped.
- Nothing is ever deleted. A wrong call stays on The Board forever with its loss attached.
Why "versus the index" is the whole game
Any bull market makes every stock picker look smart against zero. Measuring against the index strips that flattery out: it answers the only real question, would you have been better off buying the boring ETF? Most published track records fail precisely this test, which is why most published track records do not include a benchmark.
The same razor for everyone
The Pundit Scoreboard runs on identical rules — same entry logic, same benchmark, same daily marks. When a TV guest's stock pick airs, it gets logged with a link to the source, and from that day the market grades it, not the pundit's next appearance. Our automated signals get the same treatment on the Proof of Signal page: tracked forward from the day they fire, never backfilled.
The reasoning is simple: a track record that only shows the wins is not a track record, it is marketing. Publishing the misses is what makes the wins worth anything. Before trusting any number on this site — including these — check the same three things this methodology promises: the call date, the entry price against that date's actual close, and the benchmark it is measured against.
If you find a scoring hole — a call that looks backdated, an entry price that does not match the close — email us. Finding one would make the site better, and the audit trail exists so you can look.
Opinion and personal record, not investment advice.