People who run ball clubs, they think in terms of buying players. Your goal shouldn't be to buy players, your goal should be to buy wins.And in order to buy wins, you need to buy runs.

It's about getting things into the **currency of points, ** describing the game using a** single language.**

We have stats for all parts of the game. Attack efficiency, passer ratings, ace / error metrics, and don't get me started on digs, blocks, and assists per set. We lack a common language.

We cannot combine the outputs of an outside hitter's reception and attacking, because we describe their value using terms only relevant to the skill in question. I can't add Passer Rating + Attack Eff and get an output of any value.

### Enter, Expected Value.

#### The concept is simple: given the situation, what do we expect to happen?

Example - opponent sends over a freeball and your team shanks it, the coach freaks out because we expect that ball to be played very well / perfectly

Why do we expect a good result? Because we have seen this play many times, we know it's easy, and so we expect our team to execute very well here.

### Expected Value = likelihood to win the point, given the current situation

Simple examples:

- Service Ace = 100% chance to win the point
- Reception Error = 0% chance
- Setter doubles the ball = 0% chance

Other examples:

**Attacker**gets a set from zone 6 and faces a triple block = maybe**55% chance**to win the point from this position?

**Attacker**gets a set from a perfect pass, with a single blocker = maybe**75% chance**here?

**Libero**is facing an attacker with only a single blocker = the inverse of ^above^ =**25% chance**his team wins the point from this position

Taking into consideration the team's likelihood to win the rally from any given situation, we can start talking about all skills using the unified **language of points**.

### Wait, so then what is Expected Value Added?!

Expected Value Added, Expected Points Added, Points Over Expectation, whatever you want to call it... it's the same concept.

The idea is: **how much better did the player perform, relative to what we expected?**

In the case of the shanked freeball, we see the freeball coming and know we're in a good spot in the rally

Maybe we have a **70% chance** to win when receiving a freeball.

This is our **expectation**. 70% chance to win, given the situation we are dealt (receiving a freeball).

And then disaster strikes! Two players collide while going for the ball and we shank it and it hits the floor.

We now have a **0% chance** to win the rally, given the situation.

This is our **result**.

### Expected Value Added = Result - Expectation

How likely were you to win the point before the touch?

How likely were you to win the point after the touch?

The difference is Expected Value Added.

Conceptually, this is not a new idea. If you're curious for more, personally I like Kirk Goldsberry's article, DataBall, which talks at length about this in the context of the NBA. But this is a common metric across all sports and you can find articles talking about Expected Goals in soccer, Rushing Yards Over Expected in the NFL, and many many others.

This is cool and all...

### But why do I care?

Using stats the way we read them, we'll find value in players that no one else can see.

We care because if we better **understand value**, we can better **understand points**.

And if we can better **understand points**, we can **better understand winning**.

We care:

Because 50% of attacks don't score yet we assume 0 value.

Because we still use passer ratings that don't translate to reality.

Because we can't properly evaluate until we speak the same language.

Here are some quick posts showcasing how athletes add or detract value

The sooner we can focus on what's important,

**the sooner we can win more often.**