Since February we’ve been running a season-long roe vs whole urchin bait comparison: every time we fished urchin, one rod carried a roe bait and the other carried a whole or cracked urchin, and we kept a scorecard. Nineteen sessions later we have numbers. We also have a list of ways the method was flawed, which we’ll get to, because the flaws are half the point.
The question came out of our own rigging guide, which honestly hedges: roe for scent, whole urchin for durability and size selection. Hedging is fine in a guide. We wanted to know what our own scorecard would say.
How we ran it
Two rods fished simultaneously, same distance, same general spot, swapped physically every 30 minutes so neither presentation owned the better piece of bottom all day. The roe rod fished sacs of fresh roe, tied or bound as our roe preservation guide shows; the whole rod fished a banded whole urchin or a cracked half, matched roughly to the target species. We tallied bites (a distinct take, not every nibble), hookups, fish landed, fish over 3 lb, and how many casts each bait survived.
Species mix: mostly sheepshead on the Gulf side, plus our UK correspondent’s wrasse sessions, plus whatever else turned up. That mix is itself a confound. Noted.
The scorecard
| Metric | Roe baits | Whole / cracked |
|---|---|---|
| Sessions | 19 | 19 |
| Bites recorded | 214 | 131 |
| Hookups | 61 | 52 |
| Fish landed | 44 | 41 |
| Fish over 3 lb | 6 | 14 |
| Casts per bait (avg) | 1.4 | 3.6 |
Read the ratios, not the totals. Roe drew 63% more bites but converted only 28% of them to hookups; whole urchin converted 40%. Roe baits were basically single-use — 1.4 casts per bait, and plenty didn’t survive the cast itself. And the size column is the loudest number on the sheet: fish over 3 lb ate the whole urchin better than two-to-one.
What we think is going on
Roe is scent with nothing guarding it. Everything in the neighborhood finds it fast, including everything small — pinfish, juvenile wrasse, one memorable afternoon of pufferfish that we’ve written up separately. Lots of bites, most of them from fish that could inhale the sac and miss the hook, or fish we didn’t want anyway.
A whole banded urchin is a filter. Small fish peck at it and give up; the fish that commits is one with the jaws to crush it, which usually means the fish you came for. Fewer bites, better bites, and a bait still on the hook three casts later. If the season had a headline, that’s it: roe finds fish, whole urchin selects them.
Where the method leaked
Honesty section. First, “bites” is a judgment call, and the person watching the roe rod knew it was the roe rod. There’s no blinding a bait experiment; assume some counting bias toward the result we expected. Second, conditions confounded everything — the whole-urchin size advantage is partly a calendar artifact, because our biggest fish came during the February–March pre-spawn window when we happened to fish whole baits harder. Third, two anglers on two continents with different species is not a controlled trial; it’s a shared diary with columns. Fourth, we changed hook patterns mid-season on the roe rod (smaller, finer wire) and hookups improved noticeably afterward, which quietly polluted the conversion numbers for the whole spring.
We’re publishing the table anyway, because a messy tally beats a confident guess, and because nobody else appears to have counted at all.
What we’d change next season
Fixed hook patterns from session one. A per-species scorecard instead of one pooled table. A stricter bite definition — probably “rod tip moves twice within five seconds” — so both of us count the same event. And a third column we wish we’d tracked from the start: time to first bite, which we suspect is roe’s real advantage and we can only vouch for anecdotally.
The practical verdict
Fishing a new spot, or a slow day when you need to know if anything’s home: roe, no question — it’s a survey tool you can cast. Fishing a known mark for a good fish: whole or cracked urchin, and accept the quiet stretches. And on a two-rod day, do what the experiment accidentally proved works: run both, let the roe rod ring the dinner bell, and let the whole urchin sit there waiting for the one guest with proper table manners.