# Costcodle strategy Make each guess do useful work Think of the game as range finding. Start with a useful estimate, read the feedback, and spend later guesses where they remove the most uncertainty. ## Binary Search Method When feedback gives you credible lower and upper bounds, choose a midpoint between them. If the midpoint is high, move the upper bound down; if it is low, move the lower bound up. This method needs real bounds from feedback and cannot guarantee a win. ## Estimate from Package Read the product name, count, weight, and size before choosing a number. A multipack or bulk format should be compared with a similar package, not a single unit. A rounded opening estimate is useful because its direction teaches you where to look next. ## Unit Price Math For a multi-item package, divide a familiar total estimate by the number of units to sanity-check your guess. Keep the package total as your game guess; unit math is only a way to catch an implausible starting point. ## Worked Example For a $12.99 target, try $20.00, then $10.00, then $15.00, then $13.00 as feedback points between high and low. The fourth guess wins because $13.00 is inside the $12.35–$13.63 winning range. ## Four Guesses for $12.99 - $20.00High; move down. - $10.00Low; move up. - $15.00High and closer; take a smaller step. - $13.00Inside the winning range. ## How Costco Prices Behave Separate two questions: what a warehouse price tag means, and which cents occur in this game's saved bank. The answers overlap less than you might expect. Shopping knowledge can help your first estimate, but the game's direction and distance feedback should decide your next move. ### What .99 and .49 Tell You in This Puzzle Bank We counted the cents in the 3399 records shipped with this site. There are 2312 prices ending in .99 and 1069 ending in .49. Together they account for 3381 records. That makes both endings reasonable candidates when you already have a narrow dollar range. It does not make either ending a safe opening guess: $9.99 is still far from a $19.99 target. This is a count of bank entries, not unique products or a representative survey of Costco stores. The entries can repeat similar products, and the bank contains historical prices. It also has other endings. Use this pattern as a small tie-breaker, never as a reason to ignore an arrow. ### The .97 Markdown and the .00 Warehouse Markdown Costco's [2018 employee orientation guide, page 19](https://www.costco.com/wcsstore/CostcoUSBCCatalogAssetStore/Attachment/Orientation_Magazine_2018.pdf#page=19) identifies .97 as a buyer markdown and .00 as a warehouse markdown. That source explains the terms; it does not tell you today's discount, whether a particular item is available, or how much cheaper it is than another store. The current puzzle bank has no prices ending in .97 or .00. Consequently, choosing .97 because you know a store pricing tip is not a useful default in this version of Costcodle. If the bank changes, this observation will need checking again. ### What About .69 and Other Endings? Only three bank entries end in .69. We have not established an official universal meaning for that ending. Avoid treating a claimed price code as a discount guarantee. In the game, a target ending in .69 follows exactly the same 5% winning rule as any other target. ## Turn Close Feedback into a Smaller Interval The amber category contains more information than a direction alone. A close guess is within 25% of the target. For a $10 guess, that condition places the target between $8 and about $13.33. If the arrow also says too low, keep only the part above $10. That is a much smaller search area than every price above $10. Likewise, a $15 guess that is close and too high places the target at least at $12 and below $15. Combine it with the earlier $10 result and your remaining interval is roughly $12 to $13.33. A $13 guess now has a clear reason behind it. The upper and lower limits come from the 25% rule; they are not a promise that the next number will win. Do this arithmetic only when it helps. You can still play by noting a simple low/high bracket. The main habit is to carry every valid clue forward instead of treating each attempt as a fresh guess. ## Plan the Last Two Guesses Before attempt four, identify what remains uncertain. If you have only a low result, get a credible upper bound rather than making another tiny increase. If you already have both bounds, use their midpoint or a plausible package price inside the interval. The final turn should follow the evidence, not a remembered price from a different package. Hints can rescue a round with weak starting information. The broad range helps when you are searching the wrong scale; the one-dollar range helps when you are close but have little confidence left. A hint-assisted finish is recorded as assisted, so decide whether you want to keep experimenting or learn the price range directly. ## Review One Detail After the Result When the result dialog opens, check the product name again. Did you miss the count, compare fresh produce with a frozen bag, or assume a premium brand cost twice as much? Pick the specific mistake and carry that lesson into tomorrow. The [saved answer archive](/archive/) is useful for reviewing package differences, while the [Costcodle homepage](/) has a small table of historical example prices. ## Common Mistakes - Ignoring package size and count. - Moving only a few cents after far feedback. - Chasing exact cents when a 5% band is enough. - Repeating a guess without using its arrow. ## FAQ **Should my first guess be precise?** No. A rounded number that gives useful direction is often a better opener. **When should I use a midpoint?** Use one after feedback gives you credible bounds on both sides. **Does unit price decide the answer?** No. It is a comparison aid; the challenge target is the package price. **How close must I be?** Within 5%. A $12.99 target accepts $12.35–$13.63. Practice the method in the [daily game](/), check [the rules](/how-to-play/), or browse [saved dates](/archive/).