起源故事 · 高德拉特与《关键链》
Origin Story · Goldratt and Critical Chain
以色列物理学家 艾利·高德拉特 写完《目标》后追问一个问题:为什么塞了那么多安全时间,项目还是延期?
1997 年他在小说《关键链》里给出答案 —— 安全时间没有被浪费在哪一处,而是被三种系统性行为悄悄吞掉:
人们不到截止日不动手(学生综合症),任务总会膨胀到填满工期(帕金森定律),同时做多件事反而每件都慢(坏多任务)。
他的结论是:不要怪人,要修系统 —— 这正是后来关键链项目管理(CCPM)的起点。
After writing The Goal, Israeli physicist Eli Goldratt kept chewing on one question: why do projects still slip even when we stuff them with safety time?
His 1997 novel Critical Chain gave the answer — the safety time isn't lost in one place. It's quietly devoured by three systemic behaviors:
people don't start until the deadline looms (student syndrome), the work bloats to fill whatever window it's given (Parkinson's law), and juggling several tasks at once makes each one slower (bad multitasking).
His conclusion: don't blame the people — fix the system. That's the seed of Critical Chain Project Management (CCPM).
1 三个任务,被三大病一起拖垮 Three Tasks, Dragged Down by All Three Diseases
健康排程Healthy schedule2 单任务接力 vs 多任务并行 Single-Task Relay vs Multitasked Juggle
上排单任务接力:做完 A 再做 B、C,每件最早交付。下排坏多任务:三件穿插着做,看似都在推进,实则三件全都拖到最后才完成,总工期还更长。 Top: single-task relay — finish A, hand off, then B, then C. Every deliverable lands as early as possible. Bottom: bad multitasking — slice all three together. Looks busy, but nothing ships until the very end, and the total schedule is longer.
3 现实里的三大病 The Three Diseases in the Real World
延期根因:项目总延期,根因往往不是估时不准,而是安全时间被三大病系统性吞掉,再多缓冲也填不满。
Root cause of slip: chronic slippage usually isn't bad estimates — it's the three diseases systemically eating the safety time. No amount of extra buffer ever fills the hole.
单任务专注:让资源一次只全力做一件事,做完再交棒。砍掉切换损耗,每件任务都更早交付。
Single-task focus: have each resource go all-in on one task, finish it, then pass the baton. Cut the switch cost and every task ships earlier.
鼓励早交付:任务提前完成立即交棒、不藏时间,把省下的余量汇给后面真正需要的人。
Reward early handoff: finish early, hand off immediately, don't hoard the leftover days — push that slack downstream to whoever actually needs it.
CCPM 对策:关键链法用激进估时 + 集中缓冲 + 接力棒预警,从制度上拔掉三大病的根。
The CCPM antidote: aggressive estimates + pooled buffer + relay-baton signaling — Critical Chain pulls the three diseases out by the roots at the system level.
一句话In One Line
三大病的共同点是:它们都把藏在任务里的安全时间悄悄烧掉,却不让你看见。
学生综合症在开头烧,帕金森定律在中段烧,坏多任务在切换处烧。
所以单靠"多给点缓冲"永远治不好 —— 真正的解法是让任务一次只做一件、提前完成就交棒、把分散的安全时间集中到末端统一管理。
这就是下一页 50/50 估时与项目缓冲要解决的事。
All three diseases share a trick: they quietly burn the safety time hidden inside each task while keeping it invisible.
Student syndrome burns it at the front. Parkinson's law burns it in the middle. Bad multitasking burns it at every context switch.
That's why "just add more buffer" never cures the slip — the real fix is single-tasking, instant handoff on early finish, and pooling the scattered safety time into one buffer at the end of the project.
That's exactly what 50/50 estimating and the project buffer (next page) are built to do.
常见误用Common Mistakes
把延期归咎于个人不努力。三大病是排程系统的产物,要修系统而非批评人。
Blaming slippage on "lazy" team members. The three diseases are a product of the scheduling system — fix the system, don't blame the people.
用"多分配几个并行任务"来提效率。并行只会放大切换损耗,单任务专注才更快交付。
Piling on more parallel work to "boost utilization". Parallelism just multiplies the switch cost — single-task focus delivers faster.
靠层层加安全时间防延期。安全时间只会被三大病吃掉,应集中到末端做项目缓冲。
Stacking safety time on every task to prevent slip. The three diseases eat it anyway — pool it at the end as one project buffer instead.