How the Hot Dogs Strategy Actually Works

What is the Hot Dogs strategy?

One filter. One rule. Fewer, better bets.

Hot Dogs is the simplest strategy on ratethat.dog, and one of the most effective. The rule is this: find a dog with a composite score of 60 or above that has no obvious rival — meaning it's the only dog in the race above 60, and it's ahead of the next-best runner by a meaningful margin. When those two conditions are met, it surfaces on the Hot Dogs page.

That's it. There's no second filter, no grade restriction, no trap requirement. The composite score is doing all the work — it blends form, speed, and suitability into a single 0-100 number that's field-relative. When a dog scores 60+ in a race where nobody else comes close, the model is saying something clear: this runner genuinely stands out. Hot Dogs just listens.

Why 60? What the threshold means

The composite score is calibrated so that the average runner in an average race sits around 40-50. A score of 60 doesn't mean the dog wins 60% of the time — it means it's significantly above the field average in a distance-aware blend of form and fitness signals. In practice, 60+ is a relatively high bar. Most races don't produce a Hot Dog. On a busy UK card with 80-100 races, you might see 8-15 qualifying picks. Sometimes fewer.

The threshold was validated on the ultra-composite scale introduced in March 2026 — the version of the rating that uses distance-aware blends derived from a grid search over 188,000 pre-race snapshots. At 60+, the Hot Dogs strike rate is **28.3%**. On the same dataset, blind-backing every top composite pick wins at 25.5%. The three-point lift comes entirely from the 'no obvious rival' filter: it removes the races where the top dog scores well but another dog is close behind — the precisely the scenario where confidence should drop.

The 'no obvious rival' rule — why it matters

This is the part that separates Hot Dogs from a simple top-pick strategy. Greyhound races are decided by small margins — a dog rated 62 in a race where the next runner is 61 is not a Hot Dog. The gap matters as much as the absolute score.

What we're filtering for is conviction. A 63 in a race where everyone else is below 45 is a statement. The model has looked at form, speed, suitability for today's conditions, first bend rating, and distance-specific signals — and landed significantly above the field. In those races, the track record is good. In the close-scored races, the variance is high and the edge is thin. Hot Dogs only shows you the ones worth paying attention to.

How to bet Hot Dogs

Level stakes. One unit per selection. That's the recommended approach — and it's the approach the published strike rate is based on. The per-pick SP varies, but the strategy is designed as a flat-staking volume play: you're not trying to land a big price every time, you're trusting that a 28%+ strike rate at average odds around 3.5-4.5 SP produces a positive return over time.

Check the page early — ideally before 10am for the morning card and before 1pm for the afternoon. The picks appear as soon as the racecards load and predictions are generated. Early in the day you'll sometimes see a pick and then watch it disappear if a dog is withdrawn and the field balance shifts. That's not a bug — it's the system doing its job.

You don't need to bet every single Hot Dog. Some users treat it as a filter and then apply a quick sense-check — is the dog a firm market favourite at 1.8 SP? The edge at short prices is compressed. Is it running in an OR trial? Trial grades have thin form data. Hot Dogs narrows the field to the interesting dogs; your judgment on whether the price is right is still part of the process.

How many picks should I expect each day?

Fewer than you might think. On a typical busy UK racing day — 80 to 100 races across all tracks — Hot Dogs surfaces roughly 8 to 15 picks. On quieter days, 4 to 8. The strategy is deliberately selective. If you're seeing 30+ Hot Dogs in a day, something's off.

The selectivity is a feature, not a limitation. High-volume tipping strategies tend to regress to the mean — the more picks you generate, the closer your strike rate gets to random. Hot Dogs trades volume for precision. If you're staking one unit per pick and betting 10 picks a day, your daily exposure is ten units. That's manageable, and it keeps the strategy from turning into a churn.

What Hot Dogs isn't

It's not a guarantee, and it shouldn't be treated as one. A 28.3% strike rate means 71.7% of Hot Dogs lose. You will have bad runs. A stretch of six or seven losers in a row is statistically normal within that distribution and doesn't mean the strategy has broken. The only way to know whether a strategy is working is to track it over hundreds of bets, not dozens.

It's also not the full picture of the site's prediction capability. Hot Dogs is a single filter on the composite score — it doesn't use the dog selector's 30+ filter dimensions, or the ML systems' feature space, or the deep reasoning previews. Think of it as the headline pick, not the full analysis. For the fuller picture on any Hot Dog, click through to the race page and look at the form breakdown, speed ratings, and suitability scores. The context is always there.