First League · Bulgaria · Regular Season - 3

CSKA 1948–Arda Kardzhali

2–1

HT 1–1

Kick-off: Friday, 31 July 2026 at 18:15
Status: Match Finished

What the model said, and what happened

Prediction correct

This verdict goes into the public record unchanged; we do not filter out the ones we get wrong. The model's track record

Goals and red cards

6'CSKA 1948M. Diallo (assist: G. Rusev)
22'Arda KardzhaliB. Karagaren
57'CSKA 1948M. Diallo (assist: D. Grivic)
67'CSKA 1948A. Hoffmann · red card

What we expected, and what happened

Expected goals before kick-off: CSKA 1948 1.36 · Arda Kardzhali 0.88. On the pitch: 2-1.

Referee

D. Dimitrov→

4.12

yellows per match

0.09

reds per match

matches of data

Model prediction

Why this prediction?

The model puts the most likely outcome at 47.5%: “CSKA 1948 to win”. Second comes “a draw” at 29.6%.
Behind that distribution are the expected goals: 1.36 for CSKA 1948, 0.88 for Arda Kardzhali.
The highest double chance value is 77.0% (1X — CSKA 1948 win or draw).
These values were computed from 1,028 past matches.

This is a prediction, not a guarantee; a probability is not a certainty.

These sentences are built by a fixed template; every number in them comes straight from the model's computation, none invented while the text was written.

CSKA 1948 to win47.5%

Draw29.6%

Arda Kardzhali to win23.0%

When the model has said 45–50% in the past, 2,070 of 4,393 predictions were correct (47.1%). Accuracy record →

First half

Actual: 1–1

CSKA 1948 lead at half-time32.7%

Level at half-time44.6%

Arda Kardzhali lead at half-time22.8%

First-half double chance

1X — CSKA 1948 ahead or level77.2%

12 — not level at half-time55.4%

X2 — level or Arda Kardzhali ahead67.3%

These values are not a fraction of the full-match prediction; they come from a separate model trained on the first-half scores of 1028 matches. Its edge over the base rate is markedly smaller than the full-match model's — the measured figure is on the accuracy record. Read them more cautiously. First-half goal markets are not published because they failed to beat the base rate.

Expected fouls and corners

Fouls

CSKA 1948: at least 10
Arda Kardzhali: at least 10

Corners

CSKA 1948: at least 5
Arda Kardzhali: at least 3

These figures are set to hold in at least 60% of matches; in the rest the count comes in below them. No upper bound is stated — we make no "at most" claim. Yellow cards were measured and are not published: there our model did not beat the historical average.

What these probabilities rest on

Expected goals comes out of a team's attack weighed against its opponent's defence. A strong attack is damped when it meets a strong defence — which is why comparing the ratings on their own is misleading.

How many goals CSKA 1948 score1.36

Attack: CSKA 1948 score above the league average (26% more).

Opponent's defence: Arda Kardzhali's defence is close to the league average (0.86×), so it doesn't move the expectation much.

Result: These two effects give an expectation of 1.36 goals (including home advantage).

How many goals Arda Kardzhali score0.88

Attack: Arda Kardzhali score close to the league average.

Opponent's defence: CSKA 1948 concede 16% fewer than the league average, which pulls the expectation down.

Result: These two effects give an expectation of 0.88 goals (including the away penalty).

Most likely scorelines

1–114.7%1–013.8%0–010.3%

A single scoreline is always less likely than an outcome: “home win” is the sum of 1-0, 2-0, 2-1 and dozens more. So the most likely single score can be a draw while the most likely outcome is a home win; the two do not contradict each other.

Week 3 of the season. This prediction was computed from previous seasons. Summer transfers, budget changes and squad rebuilds aren't in the ratings yet — the model can only see that a team has been rebuilt once new results come in.

We measured this: the bias is only pronounced in the opening week. In this week the model's overall accuracy does not separate from the rest of the season — that is, we found no measurable flaw. The one tendency across all opening weeks is that home wins are called slightly too often (said 43.0%, got 41.9%; n=583).

This sample is small and not statistically established, which is why we do not correct the model for it. The correction we tried (pulling probabilities toward the league average) didn't win on holdout, so it wasn't added. Read opening-week predictions with a little more caution.