KHL to NHL Translation

Understanding KHL to NHL Transitions

The Kontinental Hockey League (KHL) is one of the world's premier hockey leagues outside the NHL. Successfully projecting how KHL players will perform in the NHL requires understanding league differences, ice surface size, playing style, and historical translation patterns.

KHL Translation Factors

import pandas as pd
import numpy as np
from sklearn.linear_model import LinearRegression
from sklearn.ensemble import RandomForestRegressor

# Load KHL-to-NHL transition data
khl_transitions = pd.read_csv('khl_to_nhl_transitions.csv')

# Calculate KHL translation factor
khl_players = khl_transitions[['khl_ppg', 'nhl_ppg']].dropna()

# Linear regression for base translation
model = LinearRegression()
X = khl_players[['khl_ppg']]
y = khl_players['nhl_ppg']

model.fit(X, y)
base_translation = model.coef_[0]

print(f"=== KHL to NHL Base Translation Factor ===")
print(f"Factor: {base_translation:.3f}")
print(f"R-squared: {model.score(X, y):.3f}")

# Age-adjusted translation
def calculate_age_adjusted_translation(khl_ppg, age):
    """Adjust translation factor based on player age"""
    base_nhle = khl_ppg * base_translation

    # Younger players typically translate better
    if age <= 23:
        age_factor = 1.10
    elif age <= 26:
        age_factor = 1.05
    elif age <= 28:
        age_factor = 1.00
    else:
        age_factor = 0.95

    return base_nhle * age_factor

# Position-specific translation
position_translations = khl_transitions.groupby('position').apply(
    lambda x: LinearRegression().fit(
        x[['khl_ppg']], x['nhl_ppg']
    ).coef_[0]
)

print("\n=== Position-Specific Translation Factors ===")
for position, factor in position_translations.items():
    print(f"{position}: {factor:.3f}")

# Advanced prediction model with multiple features
features = ['khl_ppg', 'age', 'height_cm', 'weight_kg',
            'khl_seasons', 'es_points_pct']

X_advanced = khl_transitions[features].dropna()
y_advanced = khl_transitions.loc[X_advanced.index, 'nhl_ppg']

# Random Forest for complex patterns
rf_model = RandomForestRegressor(
    n_estimators=100,
    max_depth=8,
    random_state=42
)

rf_model.fit(X_advanced, y_advanced)

# Feature importance
feature_importance = pd.DataFrame({
    'feature': features,
    'importance': rf_model.feature_importances_
}).sort_values('importance', ascending=False)

print("\n=== Feature Importance for KHL-NHL Translation ===")
print(feature_importance)

# Success probability analysis
def calculate_nhl_success_probability(khl_stats):
    """Calculate probability of NHL success"""
    score = 0

    # Production component (40 points)
    if khl_stats['khl_ppg'] >= 0.70:
        score += 40
    elif khl_stats['khl_ppg'] >= 0.55:
        score += 30
    elif khl_stats['khl_ppg'] >= 0.40:
        score += 20
    else:
        score += 10

    # Age component (20 points)
    if khl_stats['age'] <= 25:
        score += 20
    elif khl_stats['age'] <= 28:
        score += 15
    else:
        score += 10

    # Even-strength production (20 points)
    es_pct = khl_stats['es_points_pct']
    if es_pct >= 0.65:
        score += 20
    elif es_pct >= 0.55:
        score += 15
    else:
        score += 10

    # Experience (10 points)
    if khl_stats['khl_seasons'] >= 3:
        score += 10
    elif khl_stats['khl_seasons'] >= 2:
        score += 7
    else:
        score += 5

    # Physical readiness (10 points)
    if khl_stats['weight_kg'] >= 90 and khl_stats['height_cm'] >= 183:
        score += 10
    elif khl_stats['weight_kg'] >= 85:
        score += 7
    else:
        score += 5

    return score

# Current KHL prospects
current_khl_prospects = pd.read_csv('current_khl_players.csv')

current_khl_prospects['success_score'] = current_khl_prospects.apply(
    calculate_nhl_success_probability, axis=1
)

# Project NHL performance
current_khl_prospects['projected_nhl_ppg'] = current_khl_prospects.apply(
    lambda x: calculate_age_adjusted_translation(x['khl_ppg'], x['age']),
    axis=1
)

# Convert to full season projections
current_khl_prospects['projected_nhl_points'] = (
    current_khl_prospects['projected_nhl_ppg'] * 82
)

print("\n=== Top KHL Prospects for NHL ===")
top_khl = current_khl_prospects.sort_values(
    'success_score', ascending=False
).head(15)

print(top_khl[[
    'name', 'age', 'position', 'khl_ppg',
    'projected_nhl_ppg', 'projected_nhl_points', 'success_score'
]])

# Historical success case studies
successful_transitions = khl_transitions[
    khl_transitions['nhl_ppg'] >= 0.60
].sort_values('nhl_ppg', ascending=False)

print("\n=== Successful KHL-NHL Transitions ===")
print(successful_transitions[[
    'name', 'age_at_transition', 'khl_ppg',
    'nhl_ppg', 'position'
]].head(10))

# Playing style analysis
def classify_khl_style(row):
    """Classify playing style from KHL stats"""
    if row['shots_per_game'] >= 3.0:
        return 'High Volume Shooter'
    elif row['assists'] / row['points'] >= 0.60:
        return 'Playmaker'
    elif row['es_points_pct'] >= 0.70:
        return 'Even-Strength Producer'
    else:
        return 'Power Play Specialist'

current_khl_prospects['playing_style'] = current_khl_prospects.apply(
    classify_khl_style, axis=1
)

print("\n=== Playing Style Distribution ===")
print(current_khl_prospects['playing_style'].value_counts())

library(tidyverse)
library(randomForest)
library(scales)

# Load KHL transition data
khl_transitions <- read_csv("khl_to_nhl_transitions.csv")

# Calculate base translation factor
base_model <- lm(nhl_ppg ~ khl_ppg, data = khl_transitions)
base_translation <- coef(base_model)[2]

cat("=== KHL to NHL Base Translation Factor ===\n")
cat(sprintf("Factor: %.3f\n", base_translation))
cat(sprintf("R-squared: %.3f\n", summary(base_model)$r.squared))

# Age-adjusted translation function
calculate_age_adjusted_translation <- function(khl_ppg, age) {
  base_nhle <- khl_ppg * base_translation

  age_factor <- case_when(
    age <= 23 ~ 1.10,
    age <= 26 ~ 1.05,
    age <= 28 ~ 1.00,
    TRUE ~ 0.95
  )

  base_nhle * age_factor
}

# Position-specific translation
position_translations <- khl_transitions %>%
  group_by(position) %>%
  summarise(
    translation_factor = coef(lm(nhl_ppg ~ khl_ppg))[2],
    n = n(),
    r_squared = summary(lm(nhl_ppg ~ khl_ppg))$r.squared
  )

cat("\n=== Position-Specific Translation Factors ===\n")
print(position_translations)

# Success probability scoring
calculate_nhl_success_probability <- function(khl_ppg, age, es_points_pct,
                                             khl_seasons, weight_kg, height_cm) {
  score <- 0

  # Production (40 points)
  score <- score + case_when(
    khl_ppg >= 0.70 ~ 40,
    khl_ppg >= 0.55 ~ 30,
    khl_ppg >= 0.40 ~ 20,
    TRUE ~ 10
  )

  # Age (20 points)
  score <- score + case_when(
    age <= 25 ~ 20,
    age <= 28 ~ 15,
    TRUE ~ 10
  )

  # Even-strength production (20 points)
  score <- score + case_when(
    es_points_pct >= 0.65 ~ 20,
    es_points_pct >= 0.55 ~ 15,
    TRUE ~ 10
  )

  # Experience (10 points)
  score <- score + case_when(
    khl_seasons >= 3 ~ 10,
    khl_seasons >= 2 ~ 7,
    TRUE ~ 5
  )

  # Physical (10 points)
  score <- score + case_when(
    weight_kg >= 90 & height_cm >= 183 ~ 10,
    weight_kg >= 85 ~ 7,
    TRUE ~ 5
  )

  return(score)
}

# Current KHL prospects
current_khl_prospects <- read_csv("current_khl_players.csv") %>%
  rowwise() %>%
  mutate(
    success_score = calculate_nhl_success_probability(
      khl_ppg, age, es_points_pct, khl_seasons, weight_kg, height_cm
    ),
    projected_nhl_ppg = calculate_age_adjusted_translation(khl_ppg, age),
    projected_nhl_points = projected_nhl_ppg * 82
  ) %>%
  ungroup()

# Top prospects
top_khl <- current_khl_prospects %>%
  arrange(desc(success_score)) %>%
  head(15)

cat("\n=== Top KHL Prospects for NHL ===\n")
print(top_khl %>%
  select(name, age, position, khl_ppg, projected_nhl_ppg,
         projected_nhl_points, success_score))

# Visualize translation relationship
ggplot(khl_transitions, aes(x = khl_ppg, y = nhl_ppg)) +
  geom_point(aes(color = age), size = 3, alpha = 0.6) +
  geom_smooth(method = "lm", se = TRUE, color = "blue") +
  scale_color_gradient(low = "green", high = "red") +
  labs(title = "KHL to NHL Performance Translation",
       subtitle = sprintf("Translation Factor: %.3f", base_translation),
       x = "KHL Points Per Game",
       y = "NHL Points Per Game",
       color = "Age at\nTransition") +
  theme_minimal()

# Success rate by production tier
success_analysis <- khl_transitions %>%
  mutate(
    khl_tier = case_when(
      khl_ppg >= 0.70 ~ "Elite (0.70+)",
      khl_ppg >= 0.55 ~ "Strong (0.55-0.70)",
      khl_ppg >= 0.40 ~ "Average (0.40-0.55)",
      TRUE ~ "Below Average (<0.40)"
    ),
    nhl_success = nhl_ppg >= 0.50
  ) %>%
  group_by(khl_tier) %>%
  summarise(
    n_players = n(),
    success_rate = mean(nhl_success),
    avg_nhl_ppg = mean(nhl_ppg),
    .groups = "drop"
  )

cat("\n=== NHL Success Rate by KHL Production ===\n")
print(success_analysis)

# Historical success stories
successful_transitions <- khl_transitions %>%
  filter(nhl_ppg >= 0.60) %>%
  arrange(desc(nhl_ppg)) %>%
  head(10)

cat("\n=== Successful KHL-NHL Transitions ===\n")
print(successful_transitions %>%
  select(name, age_at_transition, khl_ppg, nhl_ppg, position))

Challenges in KHL Translation

KHL players face unique challenges adapting to NHL hockey: smaller ice surface requires quicker decisions, more physical play, and different tactical systems. The best KHL-to-NHL transitions typically feature elite even-strength production, younger age, and strong physical readiness.