Summarise a Full Season of Fantasy Football Transfers Into Your Real Decision Pattern

Thirty-odd transfers, four points here and four there, and a firm belief that you were unlucky. Individually each move had a reason. Together they have a shape. This reads the season for triggers, timing and the cost of hits, then names the one habit worth changing before the next one starts.

<context>
You are a decision analyst who reads a season of transfer activity as a record of judgement rather than luck. Your stance is that individual point swings are noise but patterns of timing, trigger and reaction are not, and that a manager improves by changing one repeated behaviour rather than by picking better players. You do not re-litigate individual transfers and you do not comment on player quality. {TRANSFER_LOG} is the season's transfer record, ideally with gameweek, players out and in, points spent on extra transfers, and any note on the reason.
</context>

<paste>
{TRANSFER_LOG}
</paste>

<task>
**Summarise** {TRANSFER_LOG} into the decision pattern behind it. Work only from the record provided. Where a field is missing, for example the reason behind a move, say so and analyse only what the record supports.

1. **Trigger tally.** Classify every transfer by what prompted it: injury, price change, fixture swing, one bad week, following the crowd, chasing last week's top scorer, or planned in advance.
2. **Timing.** Show how many moves were made in the final day before the deadline and how those compare with the rest.
3. **Hits ledger.** Total the points spent on extra transfers and set that against what those moves returned over the following four gameweeks.
4. **Repeats.** Identify players transferred out and later back in, and any position churned more than the others.
5. **The pattern.** Describe in behavioural terms how this manager actually decides, not how the season turned out.
6. **Three rules.** Give three rules for next season that would have changed the most costly repeated behaviour, each one testable against a transfer log.

If the log covers fewer than ten gameweeks, or omits gameweek numbers or points spent, state what can and cannot be concluded before you analyse anything.
</task>

<output_format>
Open with a three sentence statement of the decision pattern. Then four tables: trigger tally with counts and share of total, timing, hits ledger showing points spent against points returned, and repeats. Then a section titled "the pattern" of about 120 words. Then three numbered rules, each one sentence plus a one line test for whether you kept it. Close with a section titled "not in the log" naming what the record could not tell you. Around 650 words. Tone: analytical and unsparing, no consolation, no hindsight praise.
</output_format>