Star Then Prune the 51-200 Watchlist: A Desk Checklist on a Normal UTC Day
Star then prune the 51-200 watchlist in one UTC pass: sort Breakout Score™, require climb and confidence before a star, then drop stale names on written rules.
Updated
Star then prune is one pass over one list, run after the UTC snapshot: sort Breakout Score™ on CoinGecko ranks 51-200, star only names that show climb and clear a confidence floor, then prune every existing star that fails your written rules. Doing both in the same sitting is the point. A star list that only grows is a collection. A list that only shrinks is a graveyard. The desk checklist below keeps it a working shortlist. Breakflare is an attention tracker, so none of this outputs entries, exits, or price targets.
Key Takeaways
- Star and prune in the same session so new candidates and old stars compete for the same limited slots.
- A star needs Breakout Score™ finalist status, climb, supportive 7d momentum, and a confidence floor. Hype alone never earns one.
- Prune on column failures you wrote down in advance, not on how you feel about the story.
- Log every star and drop with a UTC date and a reason code so next week’s you can audit the habit.
- Board first, agent second: the Trading agent should inherit the exact gates you already apply by hand, and it starts in paper mode.
What does it mean to star then prune the 51-200 watchlist?
It means treating the working list as a set of dated hypotheses with two verbs. Star says: this name earned a slot today because the columns agree. Prune says: this name lost its slot because the columns stopped agreeing. Both verbs read the same fields on the same snapshot, which is why they belong in one checklist instead of two separate rituals.
The band matters. Breakflare scores the fixed rank 51-200 universe, so a star is always a mid-cap hypothesis with room to climb toward the top 50. If you are tempted to star a top-50 name or a microcap outside the band, you have left the research object this checklist was built for.
A star is not a verdict on the project. It is a note that says “keep reading this row,” plus a review date. If you want the longer argument for why stars should stay few, read star names on the band, then prune weekly. This post is the shorter, same-day version: what the desk does in one pass, in order, without negotiating.
How do you run the checklist on a normal UTC day?
Wait for the daily snapshot, then work top to bottom. The snapshot rhythm is explained in the UTC snapshot research habit. Momentum needs history, so refreshing a screen at random hours adds noise, not information. On a normal day, with no unusual market-wide event, the pass has five moves.
| Move | What you do | Stop signal |
|---|---|---|
| 1. Sort | Sort Breakout Score™ on ranks 51-200 and read only the top finalists | You are scrolling past the finalist group |
| 2. Confirm | Read hype next to 3d and 7d momentum, then climb | Hype is loud and climb is flat |
| 3. Gate | Apply your confidence floor to each remaining finalist | Coverage is thin, so the row stays exploratory |
| 4. Star | Star survivors only until the working cap is reached | The list is already at your cap |
| 5. Prune | Re-read every existing star and drop failures | A drop rule fired and you hesitate |
Run the star half first. The reason is practical: you see today’s best candidates before you decide what to cut, so the cap creates honest pressure. If you prune first, you tend to protect old names because nothing visibly competes with them. If you star first, a full list forces the question “which current star deserves this slot less than the newcomer?”
Keep the working list small enough to read every row. A common range is a handful of names, with a hard ceiling described in shortlist size limit. If you cannot name why each star is there in one line, the list is already too long.
What does one pass look like with a worked example?
Take a made-up ticker, ALPHA, to show the reasoning. These are qualitative readings, not real data.
ALPHA appears as a Breakout Score™ finalist after the snapshot. Hype is elevated versus peers. The 7d momentum column is positive and the 3d reading is not fading. Climb shows genuine rank progress inside the band, and confidence clears your floor. Every gate passes, so ALPHA gets a star with a one-line thesis: “finalist with rising 7d momentum, climb confirmed, confidence above floor.” The note carries the UTC date and a review date.
Now BETA, an existing star from last week. Hype is still loud, but 7d momentum has rolled over and climb has gone flat. Nothing about its project changed. Its columns did. The drop rule for failed momentum fires, BETA is pruned, and the log gets one line with the reason code. If you want the exact trigger language, the prune rules post lays out the automatic drops.
The lesson from the pair is that you never argued about either name. You read columns against rules. The decision took seconds, which leaves your attention for the rows that are genuinely ambiguous.
What fail modes break a star then prune checklist?
Name them out loud so they turn into filters instead of habits. Each one below shows up repeatedly in mid-cap research, and each has a specific counter.
- Starring from the loudest tab. A name trends on social, so it gets a star before the sort. The counter is mechanical: nothing is starred unless it appeared on today’s Breakout Score™ sort first.
- Starring without climb. Chatter is not rank progress. This is the pattern in high hype, no climb, and it should end in a watch note rather than a star.
- Pruning on vibes. “It felt weak today” is not a rule. If no column failed, the star stays until its review date.
- Never pruning. The cap quietly drifts upward because each keep feels small. Count the list at the start of every pass.
- Keeping for sunk effort. A long note is not a reason. The note already did its job when it taught you what to watch.
- Silent band exits. A star that climbs into the top 50 has changed research object. Switch to the after top-50 upgrade habit instead of leaving it as a mid-cap star by default.
- Skipping the confidence read on prune day. You check confidence at starring and forget it afterward, so a name whose coverage collapsed survives untouched.
Notice that most fail modes are about process, not about picking bad coins. A desk checklist cannot make a hypothesis correct. It can make your handling of hypotheses consistent, which is what makes the results readable later.
How does a dated catalyst change the pass?
Occasionally a name has a scheduled event, such as a listing or an unlock, that explains why a star should outlive a soft week. Treat that as the one legitimate exception, and handle it with an expiry, not an open-ended keep. The catalyst date versus hype timing post covers why event timing and attention timing rarely line up.
Write the exception in the log: the event, the date, and the day the star expires if the columns have not improved. When the date arrives, run the normal gates. If climb and confidence agree, the star earns a fresh review date. If they do not, it is pruned like any other name. The exception is a delay, never an immunity.
Limit yourself to a small number of these per week. A list that is mostly exceptions has stopped being a checklist and has become a collection of stories you like.
How do you turn the checklist into a written habit?
Board first, agent second. The order is deliberate, because a rule you cannot follow by hand is not ready to be automated.
Board habit. Put the five moves on a card next to your screen. Run them after each UTC snapshot you research. Keep a plain log with four fields: UTC date, ticker, action (star, keep, drop), and reason code. Reason codes can be short: finalist-climb, momentum-fail, confidence-fail, band-exit, catalyst-expiry. Short codes make weekly counting easy. If your list has no drops in a month, either the gates are too loose or you are not running the prune half.
Agent habit. The Breakflare Trading agent is included with Plus and starts in paper mode by default. Live orders use your own Binance trade-only keys and pay-as-you-go credits at 0.5% of fill notional. Encode the same gates you use on the board: a Breakout Score™ floor, a confidence skip, and a climb requirement. The score floor rules and low confidence skip posts show how those gates read in practice.
Then check the paper log against your star log. If the agent entered a name you would not have starred, the rule is incomplete. Fix the rule before you think about live keys. See paper mode first for the full reasoning.
How do you hand the list off between days?
A star list is only useful if tomorrow’s pass can pick it up cleanly. Keep one list as the source of truth and update alerts and any agent universe in the same session as the prune. If a dropped name still pings you, the list and the alerts have diverged.
For the weekly rhythm, pair this daily pass with a Friday carry and a Monday re-sort. The Friday watchlist, Monday rescore routine covers carrying names across a weekend, and the Monday reset covers rebuilding from a clean sort. The daily star then prune pass keeps both of those cheap, because the list never drifts far.
If you use live data during the day, remember that a mid-session look is not the snapshot. A post on the live board versus the UTC lock explains which vocabulary belongs to which view, so your log stays consistent.
Which checklist should you open instead?
Use star then prune when your question is “what is on my working list and should it stay?” Use a different screen when the question changes.
- If you are building the first shortlist of the day from the full band, start with the daily rank 51-200 watchlist workflow in the related reading below.
- If you want to understand why confidence is a gate at all, read the confidence score explainer in the related reading below.
- If a single row looks great on hype but not on rank, use the false-breakout checklist linked above.
- If you want the full column definitions, the scoring stack article in the related reading below is the reference.
Want live ranks, alerts, whales, and the paper Trading agent on the daily snapshot? Start a 14-day Plus trial when you are ready to run this pass on current board data. Plus is $399/mo or $3,990/yr.
One-line desk standard
On a normal UTC day, star only what climbs with confidence, prune only on rules you wrote before you looked, and log both so the habit can be audited.
Frequently asked questions
Is star then prune the 51-200 watchlist a trade signal? No. It is a research filter for organizing attention on the rank 51-200 board. Core Breakflare research does not place trades, publish price targets, or tell you when to enter or exit.
What changes on a normal UTC day? Nothing about the checklist, which is the intent. A normal day means no unusual market-wide event, so the five moves run in order. On louder days you may see more names fail climb or confidence, which means more prunes, not looser gates.
Should Trading agent use this as a hard skip? Yes, as a paper-first rule. If you would not star a name by hand, encode the same climb and confidence gates so the agent skips it too. Keep paper mode on until the paper log matches your own star log.
Does Plus change the columns involved? No. The columns and the logic are the same. Plus adds live ranks, whales, trackers, alerts, member chat, 90-day pick history, and the Trading agent, so you can run the pass on current data and sync alerts to the list.
Why star before pruning instead of the other way around? Starring first shows you today’s best candidates before you cut anything. The working cap then forces an honest comparison between a newcomer and the weakest current star, instead of defaulting to the names you already know.
How long should the pass take? Short, once the card is on your desk. Most of the time goes to the genuinely ambiguous rows. If prune sessions routinely turn into long debates, your drop rules are still negotiable and need to be rewritten more strictly.
Related reading
Start researching today's shortlist
Pay $1 for 14 days: today's live board, whales, and alerts so you can invest from research, not FOMO. Then $399/mo. Or Core at $99/mo. Cancel anytime.