BMS FPL Draft League Analysis

• H2H Pts: league table points from weekly head-to-head results, with 3 for a win and 1 for a draw.
• PF: points for, meaning the total starter points a team actually scored.
• PA: points against, meaning the total points scored by that team's weekly opponents.
• Avg: PF divided by gameweeks played.
• Top Scores: weeks where that team had the highest score in the league.
• Bottom Scores: weeks where that team had the lowest score in the league.

• A winning streak is consecutive H2H wins. A losing streak is consecutive H2H losses.
• Draws break both streak types.
• Longest Win Streak is the team's longest consecutive-win run.
• Longest Loss Streak is the team's longest consecutive-loss run.
• Win GW(s) and Loss GW(s) show up to the first three matching streak ranges; +N more means there were additional matching streaks.

• Percentages are based on completed gameweeks.
• Top 1 is the share of gameweeks where the team tied or held the highest score.
• Top 3 and Top 5 use weekly score cutoffs, so ties at the cutoff are included.
• Bottom 5, Bottom 3, and Bottom 1 work the same way from the bottom of that week's scores.
• Because tied cutoff scores are included, Top 5 and Bottom 5 can add to more than 100% for a team when tied scores straddle the middle of a week.

• xPts asks: what if every team played every other team every week?

For each gameweek, a team gets 3 all-play points for every league opponent it outscored, 1 for every tie, and 0 for every opponent that beat it. Because this is a 10-team league, each team has 9 all-play comparisons per week. I divide the season-long all-play total by 9 so xPts lands on the same scale as the real H2H table.

Example: if a team scores third-best in a week, it beat 7 teams and lost to 2, so that week contributes 21 / 9 = 2.33 xPts.

• Luck is the gap between the real H2H table and the all-play table.

Formula: Luck = Actual H2H Points - xPts.

For **Cunha Believe it?!, that is 81 - 64.8 = +16.2. The team still led the league in PF, but the schedule also broke nicely: low PA and several weeks where the opponent did not punish an ordinary score.

• The belt started with Levi Morant (*Saka It To Me Baby) going into GW1.
• Each gameweek, only the current holder's matchup matters.
• If the holder wins or draws, they keep the belt and it is up for grabs again in their next matchup.
• If the holder loses, the opponent takes the belt immediately after that gameweek.
• Number of Reigns counts every separate time a manager held the belt, including Levi's starting reign.
• Longest Defense Run is the most holder wins in a single reign before losing the belt. A draw retains the belt, but it is not counted as a defense.
• Belt Fight Record is shown as wins-losses across every lineal belt matchup a manager appeared in, whether they entered as holder or challenger.
• Avg Score in Belt Fights is that manager's average score in matchups where the lineal belt was on the line.
• Diff vs Non-Belt Avg is Avg Score in Belt Fights minus that manager's average score in all other league matchups. Positive means they scored better in belt fights than in non-belt weeks.
• In the Week-By-Week Belt Path table, each row is tinted by the manager holding the belt after that gameweek.
• Score is shown in holder-entering to challenger order. If Belt After GW matches Holder Entering, the holder defended; if it changes, the challenger won the belt.

• These are timing metrics. They do not say a team is good or bad; they say whether a weekly score met an unusually strong or weak opponent.
• Bad Beat: a team lost while scoring above that gameweek's league median.
• Soft Win: a team won while scoring below that gameweek's league median.
• Timing Balance: soft wins minus bad beats. Positive means a manager got more schedule help in these specific timing spots; negative means more painful timing.
• In the manager details, Score is that manager's weekly score, Opp Score is the opponent's weekly score, and Margin is shown from that manager's perspective.
• GW Median is the league median score for that gameweek, which is the cutoff used to tag bad beats and soft wins.

• Boom/bust is measured with the standard deviation of weekly starter scores. Higher stdev means a wider spread between explosive weeks and quiet weeks.
• Stdev: standard deviation of weekly starter points.
• Best and Worst: highest and lowest weekly scores.
• Top Scores: number of weeks where the team led the league in points.
• Bottom Scores: number of weeks where the team had the league's lowest score.

• Team-by-position scoring uses only players in the starting XI each week, so it reflects both lineup decisions and available squad quality at that position.
• Starter Points: total points scored by starters at that position.
• Starts: number of starting slots used at that position across the season.
• Points/Start: starter points divided by starts, useful because formations change how many DEF/MID/FWD slots a team uses.
• Starter Weekly Avg: average weekly score for starting lineup players at that position.
• Rostered Weekly Avg: average weekly score for every rostered player slot at that position, including starters and bench players.
• 4+ Week Pool Weekly Avg: average weekly score for players at that position who appeared on a league roster in at least 4 distinct gameweeks. Four weeks is roughly 10% of the 38-gameweek season, so it filters out one-week punts without requiring a player to be a long-term hold. Zeroes still count for qualified players.
• The full team-by-position breakdown is split into one collapsible block per position, sorted from highest Points/Start to lowest inside each position.

• This replaces the old transaction-volume metric with a stricter counterfactual question: which matchups plausibly depended on waiver or free-agent pickups after accounting for the players dropped?
• Pickup players: players acquired through an accepted waiver or free-agent transaction. Trades and drafted players are not counted as pickups for this metric.
• Pickup starter points: points scored by those pickup players only when they were in the manager's starting XI.
• Replacement Delta: pickup starter points minus the points scored that same gameweek by the player dropped to make that pickup. This handles the case where the dropped player would also have scored enough to matter.
• If a pickup is not started, it contributes 0 to this metric. The dropped player's score is only used as a counterfactual for pickup players who actually entered the manager's starting XI.
• Waiver-Inspired Win: a H2H win where the winner's replacement delta minus the opponent's replacement delta was at least the victory margin. The winner must also have a positive replacement delta.
• Waiver-Inspired Loss: the reverse view of the same matchup. If a manager loses a game that qualifies as a waiver-inspired win for the opponent, it also counts as a waiver-inspired loss for that manager.
• Share of Wins: waiver-inspired wins divided by total wins.
• Share of Losses: waiver-inspired losses divided by total losses.
• Avg Waiver-Inspired Win Margin: the average final matchup margin in that manager's waiver-inspired wins.
• Avg Waiver-Inspired Loss Margin: the average final matchup margin in that manager's waiver-inspired losses.

• Public means the endpoint returned usable data without a logged-in session.
• Public/empty without auth means the endpoint responded but did not include useful private-session data.
• Error-style access values are endpoint probes that likely require login or are not available for this entry without an authenticated session.