Connection Degree

Connection degree is the network differentiator: for any player, it tells your club how you can actually reach him — "your director of football worked with him" (first degree), or "a player you worked with was his teammate" (second degree). It is not a social graph of who-follows-whom. It is a career-overlap graph built from verifiable shared employment — people who were at the same club at the same time — so a connection means a real working relationship someone can pick up the phone about.

In one line: scouting is a relationships business. We turn the careers of your staff, the world's players, and its coaches into one graph where a link means "we were at the same club at the same time", then compute the shortest path from your club to any player — first degree (you worked with him), second degree (one trusted introduction away), or no clear path. It is a reachability map, not a friendship score.

Why shared employment, not "connections"

A social-network connection is cheap and unverifiable — clicking "connect" says nothing about whether two people ever actually worked together. Shared employment is different: if your director of football and a player were both at the same club across overlapping dates, they demonstrably shared a workplace. That is a far stronger signal of a real, callable relationship, and it comes from the career histories the platform already holds rather than from anyone's self-reported links. The price is that the graph is only ever as complete as those histories — which we're explicit about below.

How a degree is computed

The people in the graph are your club's staff (your director of football and scouts, with their own playing and coaching careers), players, and coaches. A link between two of them is a career overlap — the same club, at the same time. Connection degree is the shortest such path from your side to a target player, and it comes in three kinds:

• First degree — someone at your club worked directly with the player: they were at the same club at the same time. A direct relationship.
• Second degree, via a player — someone you worked with was, elsewhere, a teammate of the target player. "Someone you worked with played alongside him."
• Second degree, via a coach — a coach you worked with later coached at a club where the target player played. "A coach you worked with later coached him."

If neither a first- nor second-degree path exists, the answer is no clear path. Stopping at the second degree is deliberate: a third-hand introduction rarely converts, so the graph reports only the paths a recruiter could realistically use. And every path it does report is fully attributed — the person who bridges it, the club, and the years — so a badge can read "via Mark Bonner → Joe Ironside (Coventry, 2018–20)" rather than an unexplained number you have to trust.

Your network, or the club's

Connection is computed at two grains, because the right answer depends on who is asking. The per-user view shows the paths from you specifically — the load-bearing one, so each scout sees their own way in rather than a colleague's. The per-club view is the union across everyone at the club — its collective reachable network. (At a one-person recruitment operation the two are the same.) Both run exactly the same logic.

Keeping it fresh without slowing the product

Computing these paths fresh on every page load would be far too slow, so reading and updating are deliberately separated. Reads are fast and pre-computed. Whenever a career history changes — a scout completing their own past, a new colleague joining, fresh player or coach history arriving — the affected networks are flagged, and an overnight job recomputes just those and swaps in the new result. The trade is honest and stated in the product: a career edit is reflected after the overnight run, not the same second, and the contacts page tells a user their network "refreshes overnight" while that's pending.

What it powers

• Badges wherever players are listed — a first / second / no-path chip in the brand's degree colours, each with its attributed explanation on hover.
• Search and shortlists you can filter and sort by how reachable a player is, every card carrying your way in.
• Chat — ask "who in our network knows this player?" or "is there any way in to him?" and get the path back.

How we keep it honest

This is a deterministic graph computation, not a predictive model, so honesty here is about coverage and the correctness of the path, not accuracy:

• It is only as good as the careers you enter. The whole map hangs off your staff's own career histories — if a director of football omits a club he worked at, every player reachable through it goes dark. Completing that history is exactly what the contacts page is for.
• Shared employment is a proxy for a relationship. Being at the same club at the same time is strong evidence two people worked together — but the graph cannot know whether they got on. It is a reachability map, not a measure of rapport.
• Career histories are partial, especially for players from less-documented leagues, so some genuine paths are missing rather than wrong.
• Every path is auditable. Because each link is attributed to a real club and dates, you can always check why the graph claims a connection.

What it can't do

• It stops at the second degree. Only direct and one-introduction-away paths are computed; everything else is "no clear path". That is a usefulness choice — a third-hand intro rarely lands — but it is a hard cut-off, not a gentle fade.
• It depends on what's been entered. Gaps in your staff's career histories are silent blind spots until they're filled in.
• The link is shared employment, not confirmed friendship. Same club, same time is co-tenure, not a guarantee of a warm relationship.
• Coverage, not model. Players whose careers we don't hold — or hold only thinly — are harder to reach through the graph, even where a real-world path exists.

The research behind it

The design is classic network science applied to football careers.

• Milgram, S. (1967). The small-world problem. — "Degrees of separation": the shortest-path-between-people framing the whole feature rests on.
• Granovetter, M. (1973). The strength of weak ties. — Why a second-degree introduction is so valuable in practice — the network's core value, not just the direct contacts.
• Newman, M. E. J. (2001). The structure of scientific collaboration networks. — People linked by shared affiliations (here, shared clubs over overlapping dates), exactly the teammate-link construction.
• Watts, D. J. & Strogatz, S. H. (1998). Collective dynamics of 'small-world' networks. — Why short paths exist between seemingly distant people in real affiliation graphs, which is what makes a two-degree reach useful rather than sparse.