Hot Goalkeepers and BTTS: Goals Conceded per Shot Faced, Last 5
Do hot or leaky goalkeepers in recent form tell you anything about both teams scoring next time? Across 95,501 matches, the answer is: far less than the save percentage suggests.
Written by Amani TahiriBoth Teams To Score specialistReviewed by Halldóra Sigurðardóttir, Editor-in-Chief
| BTTS after both defences conceded under 20% of SOT | 51.2% |
|---|---|
| BTTS after 45%+ of SOT went in | 50.4% |
| SOT converted next match after the hot-goalkeeping spell | 25.2% |
| BTTS, all qualifying matches | 52.2% |
- Market
- BTTS
- The question
- If two defences have let in an unusually low or high share of the shots on target they faced over their last five games, does that predict BTTS in the next match?
- Sample
- 22 European divisions with match statistics, 2005/06–2025/26, 95,501 league matches
Every week I hear the same pitch: this keeper is on fire, so back a low-scoring game. Or the reverse: they're letting everything in, so goals at both ends. This study checks that habit against pre-match form, meaning goals conceded per shot on target over each side's last five league games. It does not cover in-match stats. I use it alongside my both teams to score predictions page as a filter, not a trigger.
The headline is flat. When both defences had conceded under 20% of the shots on target they faced, BTTS landed 51.2%. When 45% or more had gone in, it landed 50.4%. The overall rate across 22 divisions from 2005/06–2025/26 was 52.2%. The save-rate story pulls the eye hard and moves the result very little.
Stat line 1Hot and cold goalkeeping barely moves the BTTS rate
The dumbbell compares the market-derived BTTS figure with what actually happened in each goalkeeping band. In the under-20% band the market said 50.9% and the outcome was 51.2%. That is close enough that the market has already priced the hot spell sensibly.
The interesting end is the leaky one. In the 45%+ band the market estimate was 52.9% but BTTS landed only 50.4%. The market seems to expect the goals to keep flowing, and they don't.
View the data: btts after hot and cold goalkeeping
| Market BTTS | Actual BTTS | 95% range | |
|---|---|---|---|
| Under 20% (10,529 games) | 50.9% | 51.2% | ±1 |
| 20–24% (15,652 games) | 51.6% | 52.7% | ±0.8 |
| 25–29% (22,905 games) | 52.0% | 52.5% | ±0.6 |
| 30–34% (21,807 games) | 52.3% | 52.1% | ±0.7 |
| 35–39% (14,125 games) | 52.7% | 52.8% | ±0.8 |
| 40–44% (7,106 games) | 52.9% | 51.8% | ±1.2 |
| 45%+ (3,377 games) | 52.9% | 50.4% | ±1.7 |
| All matches (95,501 games) | 52.1% | 52.2% | ±0.3 |
The 20–24% band ran 52.7% against a market estimate of 51.6%, a gap of +1.1 points. I note it. I don't build anything on it.
Stat line 2Both extremes settle almost the same way next time
If hot keepers really strangled games, the 0-0 slice would swell in the under-20% band. It doesn't. The next match split 51.2% both scored, 40.6% one side only and 8.2% 0-0.
The 45%+ band looks almost identical: 50.4% both scored, 41.4% one side only, 8.2% 0-0. Two very different recent stories end in the same shape. That tells me the last five games are mostly noise for this market.
View the data: how the next match settled
| Both scored | One side only | 0-0 | |
|---|---|---|---|
| Under 20% | 51.2% | 40.6% | 8.2% |
| 20–24% | 52.7% | 39.7% | 7.6% |
| 25–29% | 52.5% | 40.1% | 7.4% |
| 30–34% | 52.1% | 40.6% | 7.3% |
| 35–39% | 52.8% | 39.8% | 7.4% |
| 40–44% | 51.8% | 40.6% | 7.6% |
| 45%+ | 50.4% | 41.4% | 8.2% |
Stat line 3Hot goalkeeping carries over a little into the next match
This is where the habit shows itself, faintly. After the under-20% spell, 25.2% of shots on target went in next match. After the 45%+ spell, 31.6% did. The bars climb steadily between those two ends.
So the effect isn't zero. Some of it is real keeping and some of it is the kind of shots those defences allow. Shots on target don't record quality, and a tidy save rate can simply mean the opposition kept shooting tamely from range.
View the data: shots on target that went in, next match
| Goals per shot on target in the next game, by band | |
|---|---|
| Under 20% | 25.2% |
| 20–24% | 27.9% |
| 25–29% | 29.6% |
| 30–34% | 30.6% |
| 35–39% | 31.2% |
| 40–44% | 31.4% |
| 45%+ | 31.6% |
On the touchline I learned that a keeper's form is usually the back line's form in disguise. A small carry-over in conversion rarely survives the question that actually decides BTTS: will both attacks create anything worth finishing?
Stat line 4One defence at a time, the hot spell trims a few points
Split by side, the pattern is clearer. Opponents of a home side in the under-20% band scored 64.8% next time, against 68.1% in the 45%+ band. For away defences it was 74.9% against 76.6%.
The bigger story on this chart is venue, not goalkeeping. The away line sits well above the home line in every band, so the away defence conceding is the default. The hot-keeper effect is a small adjustment on top of that.
View the data: each defence on its own
| Home defence | Away defence | |
|---|---|---|
| Under 20% | 64.8% | 74.9% |
| 20–24% | 69.1% | 76.5% |
| 25–29% | 67.7% | 77.4% |
| 30–34% | 69.3% | 77.3% |
| 35–39% | 69.0% | 77.4% |
| 40–44% | 69.2% | 77.2% |
| 45%+ | 68.1% | 76.6% |
The dip sits almost entirely in the hottest band. Everywhere else both lines run flat, so I only look at a keeper's numbers when they are genuinely extreme.
Stat line 5Fair odds barely change from the hottest band to the coldest
Fair BTTS odds ranged from 1.95 to 1.99 across the two extreme bands, which is a narrow corridor. Backed at market-fair odds, the under-20% band returned +0.6% and the 45%+ band −4.7%.
The main caveat: there are no BTTS prices in the records, so the market figure is rebuilt from match odds and Over/Under 2.5 lines. Real BTTS prices carry a margin of several points, bigger than almost every gap in this table. These returns are a benchmark against the market's view, not money you could have banked.
| Goals per SOT faced | Games | BTTS | 95% range | Fair odds | Return |
|---|---|---|---|---|---|
| Under 20% | 10,529 | 51.2% | 50.2% to 52.2% | 1.95 | +0.6% |
| 20–24% | 15,652 | 52.7% | 51.9% to 53.5% | 1.90 | +2.3% |
| 25–29% | 22,905 | 52.5% | 51.9% to 53.1% | 1.90 | +1.0% |
| 30–34% | 21,807 | 52.1% | 51.4% to 52.8% | 1.92 | −0.5% |
| 35–39% | 14,125 | 52.8% | 52.0% to 53.6% | 1.89 | +0.1% |
| 40–44% | 7,106 | 51.8% | 50.6% to 53.0% | 1.93 | −2.0% |
| 45%+ | 3,377 | 50.4% | 48.7% to 52.1% | 1.99 | −4.7% |
Most of these differences sit inside their 95% ranges. The under-20% band runs 50.2% to 52.2%, and that includes the overall rate. Read them as leans.
Sheet verdict
- Discount the save-rate headline
Hot goalkeeping over five games left BTTS at 51.2% against a 52.2% baseline. Don't let a run of saves talk you into a No on its own.
- Resist the leaky-keeper Yes
After 45%+ of shots went in, BTTS landed 50.4% while the market model expected 52.9%. Recent leaks don't promise goals at both ends.
- Check shot quality first
Next-match conversion only moved from 25.2% to 31.6% across the extremes. Before trusting a save rate, look at whether the chances conceded were clear or merely shots that got through.
Questions bettors ask
Does a goalkeeper in good form make BTTS less likely?
Only slightly. When both defences had conceded under 20% of shots on target over their last five games, BTTS landed 51.2%, against 52.2% across all matches.
Should I back BTTS when both defences have been leaking goals?
The data says be careful. In the 45%+ band BTTS landed 50.4%, and the return at market-fair odds was −4.7%, before any real bookmaker margin.
Is goals conceded per shot on target a useful BTTS stat?
It carries over a little. Conversion next match was 25.2% after hot spells and 31.6% after cold ones, but fair BTTS odds only ranged from 1.95 to 1.99.
I'm Amani Tahiri, based in Marrakesh, and I write the both-teams-to-score tips at fixedmatchesvip.com — I read where a defence cracks and whether an attack genuinely wants the goal, not which famous names are on the sheet.
View profile →
