18+ gambleaware.org

NB I Hungary Prediction Models

2026/2027

Flat Betting Model

58
Matches w/ Odds
$10
Stake Per Bet
+16.65%
Best: Under 2.5
-62.25%
Worst: Away Underdog

Odds Range Filter

1.0
10.0

Use minimum and maximum odds selection tool to identify the most profitable ranges for the Flat Betting football prediction model for NB I Hungary. After each change on the odds range, the system filters qualifying bets and recalculates profits and losses based on those selections.

Market ROI Summary

Flat $10 stake on every match from start to end of season. Profit/loss and ROI per market.

MarketBetsWinsWin %StakedReturnP&LROIBest Str.Worst Str.Trend
Home Bet582339.66%$580$502.5−$77.50-13.36%23
Draw Bet581525.86%$580$609.3+$29.30+5.05%36
Away Bet582034.48%$580$572.3−$7.70-1.33%38
Over 2.5573052.63%$570$494.4−$75.60-13.26%45
Under 2.5572747.37%$570$664.9+$94.90+16.65%54
BTTS Yes583458.62%$580$533.9−$46.10-7.95%63
Home Favorite432046.51%$430$412.3−$17.70-4.12%33
Away Underdog1218.33%$120$45.3−$74.70-62.25%111
Favorite582644.83%$580$545.1−$34.90-6.02%34
Underdog581729.31%$580$529.7−$50.30-8.67%310

Staking Plan Comparison

Same bets, different stake sizing. Flat = $10 every bet. Kelly ¼ = stake based on rolling edge estimate (quarter-Kelly for safety). Confidence = stake proportional to implied probability (more on favorites, less on longshots).

MarketBetsFlat P&LFlat ROIKelly ¼ P&LKelly ¼ ROIConf P&LConf ROIComparison
Home Bet58−$77.50-13.36%−$39.74-54.79%−$67.64-11.45%
Draw Bet58+$29.30+5.05%+$8.12+44.74%−$6.67-1.16%
Away Bet58−$7.70-1.33%−$37.06-80.93%+$8.08+1.35%
Over 2.557−$75.60-13.26%−$7.39-8.09%−$81.66-13.55%
Under 2.557+$94.90+16.65%+$0.78+1.53%+$96.97+17.1%
BTTS Yes58−$46.10-7.95%+$2.92+18.67%−$50.79-8.51%
Home Favorite43−$17.70-4.12%−$17.49-14.09%−$28.99-6.63%
Away Underdog12−$74.70-62.25%−$1.67-100%−$75.03-62.35%
Favorite58−$34.90-6.02%−$16.59-10.83%−$50.61-8.46%
Underdog58−$50.30-8.67%−$15.09-94.61%+$0.46+0.08%

Staking Methods

Flat Stake:
Fixed $10 on every qualifying bet regardless of odds or edge.
Kelly ¼:
Stake = bankroll × (edge / (odds−1)) / 4. Uses rolling win rate as probability estimate. Grows bankroll when winning, shrinks when losing.
Confidence-weighted:
Stake = (implied prob / rolling avg implied prob) × $10. Bets more on favorites, less on longshots. Uses your previous bets' average to calculate each stake — applicable in real time.

By Team Bets

Flat $10 on this team's Win result in every match. Sorted by ROI.

TeamBetsWinsWin %P&LROI
Puskás10660%+$59.30+59.3%
Győri ETO10660%+$37.80+37.8%
Debrecen10440%+$0.80+0.8%
Paks9444.44%−$1.80-2%
Újpest10440%−$8.90-8.9%
Zalaegerszeg10330%−$9.80-9.8%
Vasas9333.33%−$14.00-15.56%
Kisvárda10220%−$18.60-18.6%
Ferencváros9444.44%−$17.30-19.22%
Honvéd8225%−$26.70-33.38%
MTK11327.27%−$41.80-38%
Nyíregyháza Spartacus10220%−$44.20-44.2%

PPG Model

12
Teams
1.37
League Avg PPG
1.08
Home Factor
λ=0.85
WPPG Decay

Odds Range Filter

1.0
10.0

Use minimum and maximum odds selection tool to identify the most profitable ranges for the PPG football prediction model for NB I Hungary. After each change on the odds range, the system filters qualifying bets and recalculates profits and losses based on those selections.

PPG / WPPG Betting ROI

Backtested ROI betting $10 using incremental PPG — seeded with previous season data, blending into current-season results.

StrategyBetsWinsWin %P&LROITrend
PPG Favorite (1×2)462247.83%+$52.20+11.35%
PPG Home Pick251352%+$20.20+8.08%
PPG Away Pick21942.86%+$32.00+15.24%
WPPG Favorite (1×2)492244.9%+$26.70+5.45%
PPG Value (fair < market)501734%−$102.30-20.46%

Strategy Guide

PPG Favorite:
Backs the team with the higher PPG to win, whether home or away.
PPG Home Pick:
Bets on the home team only when PPG rates them higher than the visitor.
PPG Away Pick:
Bets on the away team only when PPG rates them higher than the home side.
WPPG Favorite:
Same as PPG Favorite but uses Weighted PPG (recent form weighted more heavily).
PPG Value:
Only bets when the PPG fair odds are lower than the bookmaker's odds — positive expected value. E.g. fair odds 1.80 vs bookmaker 2.10 = value bet.

Staking Plan Comparison

Same PPG bets, different stake sizing. Flat = $10 every bet. Kelly ¼ = stake based on rolling edge estimate. Confidence = stake proportional to implied probability.

StrategyBetsFlat P&LFlat ROIKelly ¼ P&LKelly ¼ ROIConf P&LConf ROIComparison
PPG Favorite (1×2)46+$52.20+11.35%+$22.53+28.96%+$31.52+6.45%
PPG Home Pick25+$20.20+8.08%−$0.13-0.36%+$7.50+2.74%
PPG Away Pick21+$32.00+15.24%+$9.58+34.77%+$19.71+9.13%
WPPG Favorite (1×2)49+$26.70+5.45%+$17.18+43.08%+$8.40+1.6%
PPG Value (fair < market)50−$102.30-20.46%−$29.16-89.23%−$55.43-10.97%

Staking Methods

Flat Stake:
Fixed $10 on every qualifying bet regardless of odds or edge.
Kelly ¼:
Stake = bankroll × (edge / (odds−1)) / 4. Uses rolling win rate as probability estimate.
Confidence-weighted:
Stake = (implied prob / rolling avg implied prob) × $10. Bets more on favorites, less on longshots. Uses previous bets' average — applicable in real time.

Recent Results — PPG Predictions vs Actual

How PPG predictions compared to actual match results

DateMatchPPG PickProbFair OddsResult✓
25 OctFerencváros vs MTKFerencváros54.12%1.855 – 0✓
25 OctPuskás vs Nyíregyháza SpartacusPuskás78.49%1.271 – 2✗
25 OctVasas vs PaksVasas50.1%20 – 2✗
24 OctZalaegerszeg vs KisvárdaZalaegerszeg39.73%2.521 – 1✗
24 OctGyőri ETO vs ÚjpestGyőri ETO46.36%2.161 – 0✓

Upcoming Fixtures — PPG Predictions

Win probabilities for upcoming matches based on current PPG ratings

DateMatchPPG PickProbFair Odds
31 OctHonvéd vs KisvárdaHonvéd40.99%2.44
31 OctZalaegerszeg vs MTK***0%0
31 OctGyőri ETO vs Vasas
🔒

Login or register for free to see predictions data.

01 NovPuskás vs Debrecen
01 NovFerencváros vs Paks

PPG Ratings — All Teams

Points Per Game overall, home, away, and weighted (WPPG with λ=0.85 decay). Click column headers to sort.

TeamPPG ↓H-PPGA-PPGWPPGvs AvgStrength
1Győri ETO2222.16+0.63
2Ferencváros1.891.52.21.84+0.52
3Puskás1.81.22.41.77+0.43
4Debrecen1.61.61.61.83+0.23
5Újpest1.5211.17+0.13
6Vasas1.41.41.41.15+0.03
7Paks1.41.41.41.65+0.03
8Kisvárda1.10.61.61.06-0.27
9MTK1.091.50.60.95-0.28
10Honvéd1111.01-0.37
11Zalaegerszeg0.91200.98-0.46
12Nyíregyháza Spartacus0.81.250.50.94-0.57

PPG vs WPPG — Biggest Divergence

Teams where PPG and Weighted PPG disagree most — indicates recent form shift

Újpest
PPG: 1.5WPPG: 1.17Gap: ↓ 0.33
Recent form declining — WPPG dropping below season PPG
Vasas
PPG: 1.4WPPG: 1.15Gap: ↓ 0.25
Recent form declining — WPPG dropping below season PPG
Paks
PPG: 1.4WPPG: 1.65Gap: ↑ 0.25
Recent form better than season average — WPPG rising
Debrecen
PPG: 1.6WPPG: 1.83Gap: ↑ 0.23
Recent form better than season average — WPPG rising
Győri ETO
PPG: 2WPPG: 2.16Gap: ↑ 0.16
Recent form better than season average — WPPG rising

Worked Example — Ferencváros vs MTK

Ferencváros PPG
1.75
vs
MTK PPG
1.2
P(Ferencváros) = 1.75 / (1.75+1.2) × 1.08
= 0.593 × 1.08 = 54.12% → Fair odds 1.85
P(Draw) = 0.268 × (1 − |diff|)
= 18.38% → Fair odds 5.44
P(MTK) = 1 − 54.12% − 18.38%
= 27.5% → Fair odds 3.64

Poisson / Correct Score

Select Match

0.71
Home xG (λ)
1.75
Away xG (λ)
2.46
Total E[goals]
1.68
League Avg Home
1.28
League Avg Away

Dixon-Coles Score Matrix — Honvéd vs Kisvárda

30 Oct · Each cell shows the probability of that exact scoreline. Dixon-Coles corrected (ρ = -0.13). · Upcoming

H \ A0123456
09.9%13.6%13.1%7.7%3.4%1.2%0.3%
14.7%12.0%9.3%5.4%2.4%0.8%0.2%
22.2%3.8%3.3%1.9%0.8%0.3%0.1%
30.5%0.9%0.8%0.5%0.2%0.1%0.0%
40.1%0.2%0.1%0.1%0.0%0.0%0.0%
50.0%0.0%0.0%0.0%0.0%0.0%0.0%
60.0%0.0%0.0%0.0%0.0%0.0%0.0%
Most likely High Medium Low

Derived Markets

Home Win
13.37%
Fair odds: 7.48
Draw
25.77%
Fair odds: 3.88
Away Win
60.87%
Fair odds: 1.64
Over 0.590.05%/ Under 9.95%Fair: 1.11
Over 1.571.76%/ Under 28.24%Fair: 1.39
Over 2.544.47%/ Under 55.53%Fair: 2.25
Over 3.523.22%/ Under 76.78%Fair: 4.31
Over 4.510.16%/ Under 89.84%Fair: 9.85
BTTS Yes43.36%Fair: 2.31
BTTS No56.64%Fair: 1.77

Correct Score — Top 12

10-1
13.6%A
20-2
13.11%A
31-1
12.02%D
40-0
9.95%D
51-2
9.31%A
60-3
7.65%A
71-3
5.43%A
81-0
4.7%H
92-1
3.78%H
100-4
3.35%A
112-2
3.3%D
121-4
2.38%A

Dixon-Coles vs Plain Poisson

Where the low-score correction adds value

Plain Poisson
P(0-0) = 8.5%
P(1-0) = 6.1%
P(0-1) = 14.9%
P(1-1) = 10.6%
✗ Underestimates 0-0, 1-0, 0-1
Dixon-Coles (ρ correction)
P(0-0) = 9.9% (+1.4%)
P(1-0) = 4.7% (-1.4%)
P(0-1) = 13.6% (-1.3%)
P(1-1) = 12.0% (+1.4%)
✓ Fixes low-score underestimation
💡 Why does this matter? Draws in football are dominated by 0-0 and 1-1. Plain Poisson underprices these by 15–20%. Dixon-Coles corrects this, generating better ROI on draw markets.

Dixon-Coles Betting ROI

Backtested ROI using incremental Poisson/Dixon-Coles model — xG computed only from matches played before each bet.

StrategyBetsWinsWin %P&LROITrend
DC Value (1X2 edge > 3%)311548.39%+$108.60+35.03%
DC Home Pick (>50%)11763.64%+$38.40+34.91%
DC Away Pick (>50%)6116.67%−$40.30-67.17%
DC Draw (>28%)12433.33%+$48.20+40.17%
DC Over 2.5 Value12650%−$20.10-16.75%
DC BTTS Value4250%−$9.30-23.25%

Staking Plan Comparison

Same Dixon-Coles bets, different stake sizing. Flat = $10 every bet. Kelly ¼ and Confidence-weighted alternatives.

StrategyBetsFlat P&LFlat ROIKelly ¼ P&LKelly ¼ ROIConf P&LConf ROIComparison
DC Value (1X2 edge > 3%)31+$108.60+35.03%+$55.62+16.33%+$67.50+22.24%
DC Home Pick (>50%)11+$38.40+34.91%+$0.72+0.69%+$34.41+30.64%
DC Away Pick (>50%)6−$40.30-67.17%$0.000%−$43.79-68.97%
DC Draw (>28%)12+$48.20+40.17%+$8.77+22.33%+$36.09+30.97%
DC Over 2.5 Value12−$20.10-16.75%−$2.20-19.5%−$23.80-18.66%
DC BTTS Value4−$9.30-23.25%$0.000%−$11.25-27.08%

Staking Methods

Flat Stake:
Fixed $10 on every qualifying bet regardless of odds or edge.
Kelly ¼:
Stake = bankroll × (edge / (odds−1)) / 4. Uses rolling win rate as probability estimate. Grows bankroll when winning, shrinks when losing.
Confidence-weighted:
Stake = (implied prob / rolling avg implied prob) × $10. Bets more on favorites, less on longshots. Uses your previous bets' average to calculate each stake — applicable in real time.

Team Attack & Defense Strength

Attack/defense strength relative to league average (1.00 = average). Used to derive expected goals (xG) for each match.

TeamAtt (H)Def (H)Att (A)Def (A)H GPA GP
Debrecen1.071.091.410.8355
Paks1.191.251.251.0755
Ferencváros1.190.591.250.3645
Újpest1.190.941.090.9555
Győri ETO1.090.781.170.664
Vasas1.070.941.090.8355
MTK0.991.040.941.6765
Puskás0.951.410.940.3655
Honvéd0.711.251.171.1954
Zalaegerszeg1.430.780.131.6956
Kisvárda0.481.091.090.655
Nyíregyháza Spartacus0.60.780.781.5946

How Poisson Predictions Work

Step 1 — Expected Goals:
Home xG = Home Attack Strength × Away Defense Weakness × League Avg Home Goals. Each team's attack and defense ratings are calculated relative to the league average.
Step 2 — Poisson Formula:
P(k goals) = e⁻λ × λᵏ / k! — models the probability of scoring exactly k goals given the expected goals (λ).
Step 3 — Score Matrix:
Multiply home and away goal probabilities for every combination (0-0 through 6-6) to build a full probability grid for every possible scoreline.
Step 4 — Dixon-Coles:
Apply ρ correction to fix Poisson's underestimation of low-scoring draws (0-0, 1-1, 1-0, 0-1). This is the key improvement over basic Poisson.
Step 5 — Derive Markets:
Sum probabilities from the matrix to derive 1X2, Over/Under, BTTS, and correct score probabilities — then convert to fair odds for value identification.

Elo Rating

Current Elo Ratings

All teams ranked by Elo. Start rating: 1500. K-factor: 32. Home advantage: +65 Elo.

1Ferencváros
1554
2Győri ETO
1551
3Debrecen
1531
4Puskás
1527
5Újpest
1508
6Paks
1506
7Vasas
1491
8Kisvárda
1488
9Honvéd
1470
10MTK
1470
11Nyíregyháza Spartacus
1455
12Zalaegerszeg
1453

Elo → Win Probability Table

How Elo difference translates to win probability.

Elo DiffP(Higher wins)Fair OddsTypical Matchup
050.0%2.00Evenly matched
5057.1%1.75Mid vs mid-lower
10064.0%1.56Top-8 vs bottom-8
15070.3%1.42Top-4 vs relegation
20076.0%1.32Leader vs bottom half
25080.8%1.24Leader vs bottom 3
30084.9%1.18Ferencváros vs Zalaegerszeg

Elo Rating History — Top 6

How the top 6 teams' Elo ratings evolved across the season.

146115121564
FerencvárosGyőri ETODebrecenPuskásÚjpestPaks

Elo Ratings — All Teams

Current Elo rating, peak Elo this season, and results. Start rating: 1500.

#TeamEloPeakΔGPWDL
1Ferencváros15541554+549450
2Győri ETO15511551+5110622
3Debrecen15311531+3110442
4Puskás15271550+2710604
5Újpest15081540+810433
6Paks15061506+610424
7Vasas14911534-910424
8Kisvárda14881519-1210253
9Honvéd14701500-309234
10MTK14701515-3011335
11Nyíregyháza Spartacus14551500-4510226
12Zalaegerszeg14531501-4711317

Elo Betting ROI

Backtested ROI using incremental Elo ratings — ratings computed from matches played before each bet.

StrategyBetsWinsWin %P&LROITrend
Elo Value (edge > 3%)28725%−$57.80-20.64%
Elo Favourite471736%−$116.40-24.77%
Elo Home Pick (>55%)100%−$10.00-100%
Elo Draw (>28%)15533%+$44.30+29.53%

Staking Plan Comparison

Same Elo bets, different stake sizing. Flat = $10 every bet. Kelly ¼ and Confidence-weighted alternatives.

StrategyBetsFlat P&LFlat ROIKelly ¼ P&LKelly ¼ ROIConf P&LConf ROIComparison
Elo Value (edge > 3%)28−$57.80-20.64%−$6.98-100%−$31.72-11.07%
Elo Favourite47−$116.40-24.77%−$31.93-98.92%−$103.48-20.69%
Elo Home Pick (>55%)1−$10.00-100%$0.000%−$10.00-100%
Elo Draw (>28%)15+$44.30+29.53%+$0.68+14.26%+$40.30+27.09%

Staking Methods

Flat Stake:
Fixed $10 on every qualifying bet regardless of odds or edge.
Kelly ¼:
Stake = bankroll × (edge / (odds−1)) / 4. Uses rolling win rate as probability estimate. Grows bankroll when winning, shrinks when losing.
Confidence-weighted:
Stake = (implied prob / rolling avg implied prob) × $10. Bets more on favorites, less on longshots. Uses your previous bets' average to calculate each stake — applicable in real time.

How Elo Ratings Work

Step 1 — Starting Rating:
Every team starts at 1500. Higher Elo means a stronger team.
Step 2 — Expected Score:
E(A) = 1 / (1 + 10^(R_B − R_A) / 400). If Arsenal is rated 1600 and Wolves is 1450, Arsenal's expected score is 0.71.
Step 3 — Update After Match:
R_new = R_old + K × (S − E), where S is the actual result (1 = win, 0.5 = draw, 0 = loss) and K=32 controls sensitivity.
Step 4 — Home Advantage:
Home team gets +65 Elo boost when calculating expected scores, reflecting the statistical home advantage in football.
Step 5 — Betting:
Convert Elo difference into win/draw/loss probabilities, compare against bookmaker odds, and bet when the model spots value.

Value Scanner

No upcoming matches with odds available. Please check back later.