Valorant is a competitive 5v5 tactical FPS where teams compete in professional tournaments such as the Valorant Champions Tour (VCT).

Player performance is often summarized using a rating metric, which combines multiple aspects of gameplay into a single value.

In VCT analytics, player rating is typically modeled as a linear combination of key performance components:

\[ \text{Rating} = C_1(\text{Kill Contribution}) + C_2(\text{Death Contribution}) + C_3(\text{APR}) + C_4(\text{ADR}) + C_5(\text{Survival}) \]

This metric aims to capture overall player impact, balancing offensive output, efficiency, and survivability.

This presentation explores whether higher-rated VCT players tend to record more kills.

We use:

• hypothesis testing
• p-values
• ggplot visualizations
• an interactive plotly chart

A p-value is the probability of observing results at least as extreme as the sample data, assuming the null hypothesis is true.

A small p-value suggests that the observed difference is unlikely to be due to random chance alone.

Do higher-rated VCT players have a higher average number of kills than lower-rated players?

We split players into two groups based on the median player rating:

High Rating Low Rating

\[ H_0: \mu_{high} = \mu_{low} \]

\[ H_a: \mu_{high} > \mu_{low} \] ## Preparing the Data

##   Rating Kills       group
## 1   0.68    11  Low Rating
## 2   0.46     6  Low Rating
## 3   0.59    17  Low Rating
## 4   1.21    18 High Rating
## 5   1.02    12 High Rating
## 6   1.13    30 High Rating

Players were divided into two groups using the median of the Rating variable.

## 
##  Welch Two Sample t-test
## 
## data:  high and low
## t = 23.211, df = 14133, p-value < 2.2e-16
## alternative hypothesis: true difference in means is greater than 0
## 95 percent confidence interval:
##  17.62597      Inf
## sample estimates:
## mean of x mean of y 
##  49.41319  30.44281

This test compares the average kills of higher-rated players against lower-rated players.

This plot shows the distribution of kills for higher-rated and lower-rated players.

This density plot helps compare how kills are distributed across the two groups.

This interactive plot shows the relationship between player rating and kills.

## [1] 2.660822e-117

Decision rule:

If p-value < 0.05, reject \(H_0\) If p-value >= 0.05, fail to reject \(H_0\)

A smaller p-value means stronger evidence that higher-rated players tend to get more kills.

While statistical metrics like kills and rating provide useful insights, they do not fully capture player impact.

Some important limitations include:

• Non-impactful kills: Late-round or low-impact kills may inflate statistics without contributing to winning rounds
  
• Stat padding: Players may accumulate stats in situations that do not meaningfully affect the outcome
  
• Lack of team context: Selfless plays, such as creating space or entry fragging, may result in fewer kills but are critical for team success
  
• Untracked contributions: Communication, positioning, and strategic decisions are not reflected in numerical metrics

Therefore, statistical analysis should be interpreted alongside game context and qualitative understanding.

Based on the t-test p-value, we determine whether there is significant evidence that higher-rated VCT players have more kills on average.

This analysis shows how p-values can be used in esports analytics to evaluate player performance differences.