Why a Tipster Hit Rate Is Misleading (and What to Check Instead)

“87% win rate!” is the most popular sentence in betting marketing, and the least useful. The tipster hit rate, meaning the percentage of tips that win, is the first number most people look at and the easiest one to make look good. I’m not saying it’s always a lie. I’m saying it’s half a number, and the missing half is where the money is.

Here’s what hit rate leaves out, the one table that fixes it, and what I’d look at instead.

What a tipster hit rate does and doesn’t tell you

Hit rate, also called win rate or strike rate, is simple: winning bets divided by total bets. Sixty winners from 100 tips is 60%.

What it tells you: how often you’d have collected.

What it doesn’t tell you:

• At what price. Winning 60% at odds of 1.50 and winning 60% at odds of 2.00 are different universes.
• With what stakes. A record can hide one 10-unit loss among many 1-unit wins.
• Over how many bets. Sixty percent of 20 tips is mostly luck.
• Whether the list is complete. Deleted losers do wonders for a percentage.

The first point matters most, so let’s start there.

The breakeven hit rate for every price

Every price has a breakeven hit rate: the share of bets you must win just to end up with what you started with. It’s one divided by the decimal odds. If you use fractional or American odds, run them through the odds converter first, or read how betting odds work.

Decimal odds	American	Breakeven hit rate	In other words
1.10	−1000	90.9%	Lose more than 1 in 11 and you’re down
1.25	−400	80.0%	4 winners in 5 is only breakeven
1.50	−200	66.7%	You need 2 in 3
1.80	−125	55.6%
1.91	−110	52.4%	The classic US spread price
2.00	+100	50.0%	Evens
2.50	+150	40.0%
3.00	+200	33.3%	1 in 3 is enough
5.00	+400	20.0%
10.00	+900	10.0%	9 losers in 10 is breakeven

Now hold any advertised hit rate against this table. “80% winners” at average odds of 1.25 is a tipster who has made you exactly nothing. At 1.20, the same 80% is a steady loss.

And a tipster hitting 30% at average odds of 4.00 (breakeven 25%) is doing very well, even though following them means losing seven bets in ten, which feels awful. That feeling is exactly why marketing prefers the first kind of tipster to the second.

Three tipsters, 100 bets each

Let’s make it concrete. Three imaginary tipsters, each with 100 tips at one unit per bet:

	Tipster A	Tipster B	Tipster C
Style	Heavy favourites	Underdogs and draws	Handicaps and totals
Average odds	1.20	3.00	1.91
Hit rate	80%	38%	55%
Breakeven hit rate	83.3%	33.3%	52.4%
Winnings	80 × 0.20 = +16.0	38 × 2.00 = +76.0	55 × 0.91 = +50.05
Losses	20 × 1 = −20.0	62 × 1 = −62.0	45 × 1 = −45.0
Profit	−4.0 units	+14.0 units	+5.05 units
Yield	−4.0%	+14.0%	+5.05%

The “worst” tipster by hit rate made the most money. The “best” lost money while being right four times out of five. If you only remember one table from this guide, make it this one.

One caveat so I don’t oversell B: 100 bets at odds of 3.00 is a small sample, and a big slice of that +14 could be luck. I’ll get to that.

Hit rate vs ROI vs yield

The terms get used loosely, so here’s the plain version.

Metric	Formula	What it answers	Weakness
Hit rate	winning bets ÷ total bets	How often do tips win?	Ignores the odds completely
Profit	returns − stakes, in units	How much was made?	Depends on how big a “unit” is and how many bets there were
Yield	profit ÷ total amount staked	How much was made per unit risked?	Noisy on small samples
ROI	usually the same as yield; sometimes profit ÷ starting bankroll	Depends on the definition	Ask which definition is used

Yield is the one to anchor on, because it bakes in both how often bets win and at what price. A yield of +5% means that for every 100 units staked, 5 came back as profit.

When a service quotes “ROI”, check what’s in the denominator. “300% ROI” often means profit compared with a small starting bankroll after a very large number of bets, which sounds far more impressive than the 3% yield it may really be.

Sample size: how much of the record is luck?

Even an honest hit rate wobbles. Flip a fair coin 100 times and you won’t get exactly 50 heads. Betting records are the same, and the wobble is bigger than people expect.

This table shows the range you’d typically see, 19 times out of 20, from someone whose true hit rate is 50%:

Number of bets	Typical range of the observed hit rate	Swing
50	36% to 64%	± 14 points
100	40% to 60%	± 10 points
500	45.6% to 54.4%	± 4.4 points
1,000	46.9% to 53.1%	± 3.1 points
2,500	48% to 52%	± 2 points

Read the second row again. Take a crowd of tipsters with no skill at all, tipping at evens, and roughly one in fifteen of them will show a hit rate of 58% or better over 100 bets. That’s a +16% yield, a screenshot, and a paid Telegram channel. Meanwhile, the difference between a genuinely good bettor (say 54% at 1.91) and a breakeven one (52.4%) is under two points, which you can’t reliably see until you’re into the thousands of bets.

Longer odds make this worse, not better. At odds of 5.00, wins are rare and each one moves the profit line a lot, so records swing wildly. This is also why bankroll management matters so much: even real edges come with long losing runs.

How hit rates get inflated

Without anyone technically lying, a win percentage can be dressed up in several ways.

1. Short prices only. Tip 1.15 shots and you’ll be “right” most weeks. See the breakeven table.
2. Parlays counted leg by leg. A four-leg parlay (an accumulator, in the UK) with three correct legs is a losing bet, but some services log it as “3 of 4 correct, 75%”. Try a few combinations in the parlay calculator and you’ll see how quickly the true win chance of a multiple falls.
3. Pushes and voids counted as wins, or quietly removed only when they’d hurt.
4. Selective history. Bad months deleted, a fresh channel started, “record since relaunch”.
5. Several outcomes tipped on one match, with only the winner mentioned later.
6. Tiny samples. “9 of the last 10!” is a sentence, not a track record.
7. Odds nobody could get. The win still counts, but the profit is calculated at a price that lasted thirty seconds.

Some of this is sloppy bookkeeping. Some of it is deliberate, and at the far end it shades into the tricks I describe in how to spot a scam tipster.

What to look at instead

If a tipster, an app or a prediction site wants your trust or your subscription, here’s the order I’d check things in:

1. Yield over the full record, with total bets and total units staked shown.
2. Number of bets. Under 300, stay in “interesting, let’s see” mode.
3. Average odds, so the hit rate finally means something.
4. Whether the advised odds were available to a normal customer at a mainstream bookmaker.
5. Closing line value: did their prices beat the final market price? It’s widely seen as the best early sign of skill, and I explain it in the value betting guide.
6. Who keeps the record. Self-reported spreadsheets are weaker than a record on an independent tracking platform with timestamps.

There’s a full checklist with a worked example in how to judge a tipster. For prediction apps, the equivalent of all this is calibration, which I cover in how prediction apps work. And when I score tipsters, transparency of the record is one of the criteria.

When hit rate is actually useful

I’ve been hard on it, so in fairness, hit rate isn’t worthless. It tells you what following someone will feel like.

A 30% hit rate at long odds means regular runs of ten or more straight losers, even when the tipster is good. Plenty of people can’t stomach that and quit at the worst moment. A 60% hit rate at shorter prices is a smoother ride for the same yield. Knowing which you’re signing up for helps you choose stakes you can live with and stick with.

So use hit rate for setting expectations. Use yield, sample size and price quality for deciding whether someone is any good.

No record, however strong, makes the next bet safe, so stake only what you can afford to lose and see the responsible gambling page if it stops being fun.