How does the optimizer decide where points go?

It allocates one point at a time to the stat with the highest current marginal value. With diminishing returns, each stat's marginal value drops as you invest, so points naturally spread across stats instead of dumping into a single one.

What is a diminishing-returns factor?

It is a number from 0 to 1 that shrinks each successive point's value. A factor of 0.95 means each point is worth 95% of the previous one. A factor of 1 means no diminishing returns and the highest-weight stat would take every point.

Where do I get stat weights?

Use a simulator like Raidbots or a class guide. Stat weights express how much one point of each stat improves your output relative to others, which is exactly what this optimizer needs.

Can it model stat caps or breakpoints?

Not directly. The greedy marginal approach handles smooth diminishing returns well, but hard breakpoints — where a stat suddenly gains or loses value at a threshold — need manual adjustment of the weight at that point.

Is the greedy result always optimal?

For weights with monotonically decreasing marginal value (the diminishing-returns model here), greedy allocation is provably optimal. For non-monotonic value curves it is a strong heuristic but not guaranteed best.

World of Warcraft Stat Optimizer

Email me this result

Get this tool's output sent to your inbox, plus one useful tool a week. No spam, unsubscribe any time.

Dumping every point into your highest stat is rarely optimal once diminishing returns kick in. This optimizer allocates points one at a time to whichever stat currently gives the most output, producing the split that maximises your total effective score.

How it works

Each stat has a weight w (output per point) and a diminishing factor d between 0 and 1. After k points already invested, the next point’s value is:

marginal value = w × d^k

The optimizer repeatedly gives the next point to the stat with the highest current marginal value, then updates that stat’s k. Because marginal value only decreases, this greedy allocation is optimal for the model.

Example and tips

With three stats weighted 10, 8, and 6 and a 0.95 diminishing factor each, the first few points go to the 10-weight stat, but as its marginal value falls below 8 the optimizer switches, and the final split is balanced rather than lopsided. Set the diminishing factor to 1 to model linear stats (all points go to the top weight), and lower it toward 0.9 to model stronger soft caps. For hard breakpoints, split a stat into two entries with different weights.